experiment 3 factors

• PID home page »
• 5. Design and Analysis of Experiments »
• 5.8. Full factorial designs »
• 5.8.5. Example: design and analysis of a three-factor experiment

5.8.5. Example: design and analysis of a three-factor experiment ¶

This example should be done by yourself. It is based on Question 19 in the exercises for Chapter 5 in Box, Hunter and Hunter (2nd edition).

The data are from a plastics molding factory that must treat its waste before discharge. The \(y\) -variable represents the average amount of pollutant discharged (lb per day), while the three factors that were varied were

\(C\) = the chemical compound added (choose either chemical P or chemical Q) \(T\) = the treatment temperature (72 °F or 100 °F) \(S\) = the stirring speed (200 rpm or 400 rpm) \(y\) = the amount of pollutant discharged (lb per day) Experiment Order \(C\) \(T\) [°F] \(S\) [rpm] \(y\) [lb] 1 5 Choice P 72 200 5 2 6 Choice Q 72 200 30 3 1 Choice P 100 200 6 4 4 Choice Q 100 200 33 5 2 Choice P 72 400 4 6 7 Choice Q 72 400 3 7 3 Choice P 100 400 5 8 8 Choice Q 100 400 4

Draw a geometric figure that illustrates the data from this experiment.

Calculate the main effect for each factor by hand.

For the C effect , there are four estimates of \(C\) : \[\displaystyle \frac{(+25) + (+27) + (-1) + (-1)}{4} = \frac{50}{4} = \bf{12.5}\] For the T effect , there are four estimates of \(T\) : \[\displaystyle \frac{(+1) + (+3) + (+1) + (+1)}{4} = \frac{6}{4} = \bf{1.5}\] For the S effect , there are four estimates of \(S\) : \[\displaystyle \frac{(-27) + (-1) + (-29) + (-1)}{4} = \frac{-58}{4} = \bf{-14.5}\]

Calculate the 3 two-factor interactions (2fi) by hand, recalling that interactions are defined as the half difference going from high to low.

For the CT interaction , there are two estimates of \(CT\) . Recall that interactions are calculated as the half difference going from high to low. Consider the change in \(C\) when \(T_\text{high}\) (at \(S\) high) = \(4 - 5 = -1\) \(T_\text{low}\) (at \(S\) high) = \(3 - 4 = -1\) This gives a first estimate of \([(-1) - (-1)]/2 = 0\) . Similarly, \(T_\text{high}\) (at \(S\) low) = \(33 - 6 = +27\) \(T_\text{low}\) (at \(S\) low) = \(30 - 5 = +25\) gives a second estimate of \([(+27) - (+25)]/2 = +1\) . The average CT interaction is therefore \((0 + 1)/2 = \mathbf{0.5}\) . You can interchange \(C\) and \(T\) and still get the same result. For the CS interaction , there are two estimates of \(CS\) . Consider the change in \(C\) when \(S_\text{high}\) (at \(T\) high) = \(4 - 5 = -1\) \(S_\text{low}\) (at \(T\) high) = \(33 - 6 = +27\) This gives a first estimate of \([(-1) - (+27)]/2 = -14\) . Similarly, \(S_\text{high}\) (at \(T\) low) = \(3 - 4 = -1\) \(S_\text{low}\) (at \(T\) low) = \(30 - 5 = +25\) gives a second estimate of \([(-1) - (+25)]/2 = -13\) . The average CS interaction is therefore \((-13 - 14)/2 = \mathbf{-13.5}\) . You can interchange \(C\) and \(S\) and still get the same result. For the ST interaction , there are two estimates of \(ST\) : \((-1 + 0)/2 = \mathbf{-0.5}\) . Calculate in the same way as above.

Calculate the single three-factor interaction (3fi).

There is only a single estimate of \(CTS\) . The \(CT\) effect at high \(S\) is 0, and the \(CT\) effect at low \(S\) is \(+1\) . The \(CTS\) interaction is then \([(0) - (+1)] / 2 = \mathbf{-0.5}\) . You can also calculate this by considering the \(CS\) effect at the two levels of \(T\) , or by considering the \(ST\) effect at the two levels of \(C\) . All three approaches give the same result.

Compute the main effects and interactions using matrix algebra and a least squares model.

Use computer software to build the following model and verify that:

Learning notes:

The chemical compound could be coded either as (chemical P = \(-1\) , chemical Q = \(+1\) ) or (chemical P = \(+1\) , chemical Q = \(-1\) ). The interpretation of the \(x_C\) coefficient is the same, regardless of the coding. Just the tabulation of the raw data gives us some interpretation of the results. Why? Since the variables are manipulated independently, we can just look at the relationship of each factor to \(y\) , without considering the others. It is expected that the chemical compound and speed have a strong effect on \(y\) , but we can also see the chemical \(\times\) speed interaction. You can see this last interpretation by writing out the full \(\mathbf{X}\) design matrix and comparing the bold column, associated with the \(b_\text{CS}\) term, with the \(y\) column.

A note about magnitude of effects

In this text we quantify the effect as the change in response over half the range of the factor. For example, if the center point is 400 K, the lower level is 375 K and the upper level is 425 K, then an effect of "-5" represents a reduction in \(y\) of 5 units for every increase of 25 K in \(x\) .

We use this representation because it corresponds with the results calculated from least-squares software. Putting the matrix of \(-1\) and \(+1\) entries into the software as \(\mathbf{X}\) , along with the corresponding vector of responses, \(y\) , you can calculate these effects as \(\mathbf{b} = \left(\mathbf{X}^T\mathbf{X}\right)^{-1}\mathbf{X}\mathbf{y}\) .

Other textbooks, specifically Box, Hunter and Hunter, will report effects that are double ours. This is because they consider the effect to be the change from the lower level to the upper level (double the distance). The advantage of their representation is that binary factors (catalyst A or B; agitator on or off) can be readily interpreted, whereas in our notation, the effect is a little harder to describe (simply double it!).

The advantage of our methodology, though, is that the results calculated by hand would be the same as those from any computer software with respect to the magnitude of the coefficients and the standard errors, particularly in the case of duplicate runs and experiments with center points.

Remember: our effects are half those reported in Box, Hunter and Hunter, and in some other textbooks; our standard error would also be half of theirs. The conclusions drawn will always be the same, as long as one is consistent.

Example of Create General Full Factorial Design

A marketing manager wants to study the influence that three categorical factors have on the ability of test subjects to recall an online advertisement. Because the experiment includes factors that have 3 levels, the manager uses a general full factorial design.

• Choose Stat > DOE > Factorial > Create Factorial Design .
• Under Type of Design , select General full factorial design .
• From Number of factors , select 3 .
• Click Designs .
• Under Name , for Factor A, type Website , for Factor B, type Product , and for Factor C, type Message style .
• Under Number of Levels , enter 3 for each factor. Click OK .
• Click Factors .
• Under Type , select Text for each factor.
• Under Level Values , for Website, name the levels News , Social Media , and Sports .
• Under Level Values , for Product, name the levels Car , Video Game , and Medicine .
• Under Level Values , for Message style, name the levels You know you should., Just the facts., and That is awesome! .
• Click Results .
• Select Summary table and design table .
• Click OK in each dialog box.

Interpret the results

The first table gives a summary of the design: the total number of factors, runs, blocks, and replicates.

The design table shows the experimental conditions or settings for each of the factors for the design points using coded factor names and levels. For example, in the first run of the experiment, Factor A is at level 1. Factors B and C are at level 3. With 3 factors that each have 3 levels, the design has 27 runs. In the worksheet, Minitab displays the names of the factors and the names of the levels. Because the manager created a full factorial design, the manager can estimate all of the interactions among the factors.

Minitab randomizes the design by default, so when you create this design, the run order will not match the order in the example output.

Design Summary

Design Table (randomized)

• Minitab.com
• License Portal
• Cookie Settings

You are now leaving support.minitab.com.

Click Continue to proceed to:

Teach yourself statistics

What is a Full Factorial Experiment?

This lesson describes full factorial experiments. Specifically, the lesson answers four questions:

• What is a full factorial experiment?
• What causal effects can we test in a full factorial experiment?
• How should we interpret causal effects?
• What are the advantages and disadvantages of a full factorial experiment?

What is a Factorial Experiment?

A factorial experiment allows researchers to study the joint effect of two or more factors on a dependent variable . Factorial experiments come in two flavors: full factorials and fractional factorials. In this lesson, we will focus on the full factorial experiment, not the fractional factorial.

Full Factorial Experiment

A full factorial experiment includes a treatment group for every combination of factor levels. Therefore, the number of treatment groups is the product of factor levels. For example, consider the full factorial design shown below:

Factor A has two levels, factor B has three levels, and factor C has four levels. Therefore, this full factorial design has 2 x 3 x 4 = 24 treatment groups.

Full factorial designs can be characterized by the number of treatment levels associated with each factor, or by the number of factors in the design. Thus, the design above could be described as a 2 x 3 x 4 design (number of treatment levels) or as a three-factor design (number of factors).

Fractional Factorial Experiments

The other type of factorial experiment is a fractional factorial. Unlike full factorial experiments, which include a treatment group for every combination of factor levels, fractional factorial experiments include only a subset of possible treatment groups.

Causal Effects

A full factorial experiment allows researchers to examine two types of causal effects: main effects and interaction effects. To facilitate the discussion of these effects, we will examine results (mean scores) from three 2 x 2 factorial experiments:

Experiment I: Mean Scores

Experiment II: Mean Scores

Experiment III: Mean Scores

Main Effects

In a full factorial experiment, a main effect is the effect of one factor on a dependent variable, averaged over all levels of other factors. A two-factor factorial experiment will have two main effects; a three-factor factorial, three main effects; a four-factor factorial, four main effects; and so on.

How to Measure Main Effects

To illustrate what is going on with main effects, let's look more closely at the main effects from Experiment I:

Assuming there were an equal number of observations in each treatment group, we can compute the main effect for Factor A as shown below:

Effect of A at level B 1 = A 2 B 1 - A 1 B 1 = 2 - 5 = -3

Effect of A at level B 2 = A 2 B 2 - A 1 B 2 = 5 - 2 = +3

Main effect of A = ( -3 + 3 ) / 2 = 0

And we can compute the main effect for Factor B as shown below:

Effect of B at level A 1 = A 1 B 2 - A 1 B 1 = 5 - 2 = +3

Effect of B at level A 2 = A 2 B 2 - A 2 B 1 = 2 - 5 = -3

Main effect of B = ( 3 - 3 ) / 2 = 0

In a similar fashion, we can compute main effects for Experiment II (see Problem 1 ) and Experiment III (see Problem 2 ).

Warning: In a full factorial experiment, you should not attempt to interpret main effects until you have looked at interaction effects. With that in mind, let's look at interaction effects for Experiments I, II, and III.

Interaction Effects

In a full factorial experiment, an interaction effect exists when the effect of one independent variable depends on the level of another independent variable.

When Interactions Are Present

The presence of an interaction can often be discerned when factorial data are plotted. For example, the charts below plot mean scores from Experiment I and from Experiment II:

Experiment I

Experiment II

In Experiment I, consider how the dependent variable score is affected by level A1 versus level A2. In the presence of B1, the dependent variable score is bigger for A1 than for A2. But in the presense of B2, the reverse is true - the dependent variable score is bigger for A2 than for A1.

In Experiment II, level C1 is associated with a little bit bigger dependent variable score in the presence of D1; but a much bigger dependent variable score in the presence of D2.

In both charts, the way that one factor affects the dependent variable depends on the level of another factor. This is the definition of an interaction effect. In charts like these, the presence of an interaction is indicated by non-parallel plotted lines.

Note: These charts are called interaction plots. For guidance on creating and interpreting interaction plots, see Interaction Plots .

When Interactions Are Absent

Now, look at the chart below, which plots mean scores from Experiment III:

Experiment III

In this chart, E1 has the same effect on the dependent variable, regardless of the level of Factor F. At each level of Factor F, the dependent variable is 2 units bigger with E1 than with E2. So, in this chart, there is no interaction between Factors E and F. And you can tell at a glance that there is no interaction, because the plotted lines are parallel.

Number of Interactions

The number of interaction effects in a full factorial experiment is determined by the number of factors. A two-factor design (with factors A and B) has one two-way interaction (the AB interaction). A three-factor design (with factors A, B, and C) has one three-way interaction (the ABC interaction) and three two-way interactions (the AB, AC, and BC interactions).

A general formula for finding the number of interaction effects (NIE) in a full factorial experiment is:

where k C r is the number of combinations of k things taken r at a time, k is the number of factors in the full factorial experiment, and r is the number of factors in the interaction term.

Note: If you are unfamiliar with combinations, see Combinations and Permutations .

How to Interpret Causal Effects

Recall that the purpose of conducting a full factorial experiment is to understand the joint effects (main effects and interaction effects) of two or more independent variables on a dependent variable. When a researcher looks at actual data from an experiment, small differences in group means are expected, even when independent variables have no causal connection to the dependent variable. These small differences might be attributable to random effects of unmeasured extraneous variables .

So the real question becomes: Are observed effects significantly bigger than would be expected by chance - big enough to be attributable to a main or interaction effect rather than to an extraneous variable? One way to answer this question is with analysis of variance. Analysis of variance will test all main effects and interaction effects for statistical significance. Here is how to interpret the results of that test:

• If no effects (main effects or interaction effects) are statistically significant, conclude that the independent variables do not affect the dependent variable.
• If a main effect is statistically significant, conclude that the main effect does affect the dependent variable.
• If an interaction effect is statistically significant, conclude that the interaction factors act in combination to affect the dependent variable.

Recognize that it is possible for factors to affect the dependent variable, even when the main effects are not statistically significant. We saw an example of that in Experiment I.

In Experiment I, both main effects were zero; yet, the interaction effect is dramatic. The moral here is: Do not attempt to interpret main effects until you have looked at interaction effects.

Note: To learn how to implement analysis of variance for a full factorial experiment, see ANOVA With Full Factorial Experiments .

Advantages and Disadvantages

Analysis of variance with a full factorial experiment has advantages and disadvantages. Advantages include the following:

• The design permits a researcher to examine multiple factors in a single experiment.
• The design permits a researcher to examine all interaction effects.
• The design requires subjects to participate in only one treatment group.

Disadvantages include the following:

• When the experiment includes many factors and levels, sample size requirements may be excessive.
• The need to include all treatment combinations, regardless of importance, may waste resources.

Test Your Understanding

The table below shows results from a 2 x 2 factorial experiment.

Assuming equal sample size in each treatment group, what is the main effect for both factors?

(A) -2 (B) 3.5 (C) 4 (D) 7 (E) 14

The correct answer is (A). We can compute the main effect for Factor C as shown below:

Effect of C at level D 1 = C 2 D 1 - C 1 D 1 = 4 - 5 = -1

Effect of C at level D 2 = C 2 D 2 - C 1 D 2 = 1 - 4 = -3

Main effect of C = ( -1 + -3 ) / 2 = -2

And we can compute the main effect for Factor D as shown below:

Effect of D at level C 1 = C 1 D 2 - C 1 D 1 = 4 - 5 = -1

Effect of D at level C 2 = C 2 D 2 - C 2 D 1 = 1 - 4 = -3

Main effect of D = ( -1 + -3 ) / 2 = -2

(A) -12 (B) -2 (C) 0 (D) 3 (E) 4

The correct answer is (B). We can compute the main effect for Factor E as shown below:

Effect of E at level F 1 = E 2 F 1 - E 1 F 1 = 3 - 5 = -2

Effect of E at level F 2 = E 2 F 2 - E 1 F 2 = 1 - 3 = -2

Main effect of E = ( -2 + -2 ) / 2 = -2

And we can compute the main effect for Factor F as shown below:

Effect of F at level C 1 = E 1 F 2 - E 1 F 1 = 3 - 5 = -2

Effect of F at level C 2 = E 2 F 2 - E 2 F 1 = 1 - 3 = -2

Main effect of F = ( -2 + -2 ) / 2 = -2

Consider the interaction plot shown below. Which of the following statements are true?

(A) There is a non-zero interaction between Factors A and B. (B) There is zero interaction between Factors A and B. (C) The plot provides insufficient information to describe the interaction.

The correct answer is (B). At every level of Factor B, the difference between A1 and A2 is 3 units. Because the effect of Factor A is constant (always 3 units) at every level of Factor B, there is no interaction between Factors A and B.

Note: The parallel pattern of lines in the interaction plot indicates that the AB interaction is zero.

• They are orthogonal arrays. Some analysts believe this simplifies the analysis and interpretation of results while other analysts believe it does not.
• They obtain a lot of information about the main effects in a relatively few number of runs.
• You can test whether non-linear terms are needed in the model, at least as far as the three-level factors are concerned.
• They provide limited information about interactions.
• They require more runs than a comparable 2 k - p design, and a two-level design will often suffice when the factors are continuous and monotonic (many three-level designs are used when two-level designs would have been adequate).

User Preferences

Content preview.

Arcu felis bibendum ut tristique et egestas quis:

• Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris
• Duis aute irure dolor in reprehenderit in voluptate
• Excepteur sint occaecat cupidatat non proident

Keyboard Shortcuts

Lesson 9: 3-level and mixed-level factorials and fractional factorials, overview section  .

These designs are a generalization of the \(2^k\) designs. We will continue to talk about coded variables so we can describe designs in general terms, but in this case we will be assuming in the \(3^k\) designs that the factors are all quantitative. With \(2^k\) designs we weren't as strict about this because we could have either qualitative or quantitative factors. Most \(3^k\) designs are only useful where the factors are quantitative. With \(3^k\) designs we are moving from screening factors to analyzing them to understand what their actual response function looks like.

With 2 level designs, we had just two levels of each factor. This is fine for fitting a linear, straight line relationship. With three level of each factor we now have points at the middle so we will are able to fit curved response functions, i.e. quadratic response functions. In two dimensions with a square design space, using a \(2^k\) design we simply had corner points, which defined a square that looked like this:

In three dimensions the design region becomes a cube and with four or more factors it is a hypercube which we can't draw.

We can label the design points, similar to what we did before – see the columns on the left. However for these design we prefer the other way of coding, using {0,1,2} which is a generalization of the {0,1} coding that we used in the \(2^k\) designs. This is shown in the columns on the right in the table below:

For either method of coding, the treatment combinations represent the actual values of \(X_1\) and \(X_2\), where there is some high level, a middle level and some low level of each factor. Visually our region of experimentation or region of interest is highlighted in the figure below when \(k = 2\):

If we look at the analysis of variance for a \(k = 2\) experiment with n replicates, where we have three levels of both factors we would have the following:

Important idea used for confounding and taking fractions

How we consider three level designs will parallel what we did in two level designs, therefore we may confound the experiment in incomplete blocks or simply utilize a fraction of the design. In two-level designs, the interactions each have 1 d.f. and consist only of +/- components, so it is simple to see how to do the confounding. Things are more complicated in 3 level designs, since a p-way interaction has \(2^p\) d.f. If we want to confound a main effect (2 d.f.) with a 2-way interaction (4 d.f.) we need to partition the interaction into 2 orthogonal pieces with 2 d.f. each. Then we will confound the main effect with one of the 2 pieces. There will be 2 choices. Similarly, if we want to confound a main effect with a 3-way interaction, we need to break the interaction into 4 pieces with 2 d.f. each. Each piece of the interaction is represented by a psuedo-factor with 3 levels. The method given using the Latin squares is quite simple . There is some clever modulus arithmetic in this section, but the details are not important. The important idea is that just as with the \(2^k\)designs, we can purposefully confound to achieve designs that are efficient either because they do not use the entire set of \(3^k\)runs or because they can be run in blocks which do not disturb our ability to estimate the effects of most interest.

Following the text, for the A*B interaction, we define the pseudo factors, which are called the AB component and the \(AB^2\) component. These components could be called pseudo-interaction effects. The two components will be defined as a linear combination as follows, where \(X_1\) is the level of factor A and \(X_2\) is the level of factor B using the {0,1,2} coding system. Let the \(AB\) component be defined as

\(L_{AB}=X_{1}+X_{2}\ (mod3)\)

and the \(AB^2\) component will be defined as:

\(L_{AB^2}=X_{1}+2X_{2}\ (mod3)\)

Using these definitions we can create the pseudo-interaction components. Below you see that the AB levels are defined by \(L_{AB}\) and the \(AB^2\) levels are defined by \(L_{AB^2}\).

This table has entries {0, 1, 2} which allow us to confound a main effect or either component of the interaction A*B. Each of these main effects or pseudo interaction components have three levels and therefore 2 degrees of freedom.

This section will also discuss partitioning the interaction SS's into 1 d.f. sums of squares associated with a polynomial, however, this is just polynomial regression. This method does not seem to be readily applicable to creating interpretable confounding patterns.

• Application of \(3^k\) factorial designs, the interaction components and relative degrees of freedom
• How to perform blocking of \(3^k\) designs in \(3^p\) number of blocks and how to choose the effect(s) which should be confounded with blocks
• Concept of “Partial Confounding” in replicated blocked designs and its advantages
• How to generate reasonable \(3^{k-p}\) fractional factorial designs and understand the alias structure
• The fact that Latin square and Graeco-Latin square designs are special cases of \(3^k\) fractional  factorial design
• Mixed level factorial designs and their applications

Factorial Experiment

• First Online: 29 November 2016

Cite this chapter

• Pradip Kumar Sahu 2  

1458 Accesses

1 Citations

Basic experimental designs what we have discussed in the previous chapter take care of one type/group of treatments at a time. If an experimenter wants to test more than one type/group of treatments, then more than one set of experiments are required to be set, thereby requiring a huge amount of resources (land, money, other inputs) and time. Even with ample resources and time, desirable information may not be obtained from simple experiments. Suppose an experimenter wants to know not only the best treatment from each of the two sets of treatments but also wants to know the interaction effects of the two sets of treatments. This information cannot be obtained by conducting two separate sets of simple experiments with two groups/types of treatments. Let us suppose an experimenter wants to know (i) the best varieties among five newly developed varieties of a crop (ii) the best dose of nitrogenous fertilizer for the best yield of the same crop and (iii) also wants to know which variety among the five varieties under which dose of nitrogen provides the best yield (i.e., variety and dose interaction effect). The first two objectives (i.e., the best variety and best dose of nitrogen) can be accomplished by framing two separate simple experiments (one with five varieties and the other one with different doses of nitrogen with a single variety), but the third objective, i.e., interaction of varieties with different doses of nitrogen, cannot be obtained from these two experiments. For this purpose we are to think for an experiment which can accommodate both the groups of treatments together. Thus, in agriculture and other experiments, the response of different doses/levels of one group of treatments (factor) is supposed to vary over the different doses or levels of other set(s) of treatments (factor(s)). Factorial experiments are such a mechanism in which more than one group (factor) of treatments can be accommodated in one experiment, and from the experiment, not only the best treatment in each group of treatments could be identified but also the interaction effects among the treatments in different groups could also be estimated.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime
• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Author information

Authors and affiliations.

Department of Agricultural Statistics, Bidhan Chandra Krishi Viswavidyalaya, Mohanpur, WB, India

Pradip Kumar Sahu

You can also search for this author in PubMed   Google Scholar

Rights and permissions

Reprints and permissions

Copyright information

© 2016 Springer India

About this chapter

Sahu, P.K. (2016). Factorial Experiment. In: Applied Statistics for Agriculture, Veterinary, Fishery, Dairy and Allied Fields. Springer, New Delhi. https://doi.org/10.1007/978-81-322-2831-8_11

Download citation

DOI : https://doi.org/10.1007/978-81-322-2831-8_11

Published : 29 November 2016

Publisher Name : Springer, New Delhi

Print ISBN : 978-81-322-2829-5

Online ISBN : 978-81-322-2831-8

eBook Packages : Mathematics and Statistics Mathematics and Statistics (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Design of Experiments (DOE)

Design of Experiments (DOE) is a study of the factors that the team has determined are the key process input variables (KPIV's) that are the source of the variation or have an influence on the mean of the output.

DOE are used by marketers, continuous improvement leaders, human resources, sales managers, engineers, and many others. When applied to a product or process the result can be increased quality performance, higher yields, lower variation of outputs, quicker development time, lower costs, and increased customer satisfaction .

It is a sequence of tests where the input factors are systematically changed according to a design matrix.  The DOE study is first started by setting up an experiment with a specific number if runs with one of more factors (inputs) with each given two or more levels or settings. 

The DOE process has a significant advantage above trial and error methods. Yes, it may end up taking more time and resources but the result will most likely be more robust.

This initial time and effort up front can be costly (this is up to the team to decide how many experiments to conduct) and time consuming but the end result will be the maximizing the outputs shown above in bold. 

A DOE (or set of DOE's) will help develop a prediction equation for the process in terms of Y = f(X 1 ,X 2 ,X 3 ,X 4 ,....X n ). GOAL:

1) Understand the influential variables and understand any interactions

2) Quantify the effect of the variables on the outputs

3) Determine the setting that optimize your response (which could be to minimize or maximize an value for an output "y")

Designing a DOE

The number of runs (treatments) depends on amount of resources that can be afforded...such as time and money and keep in mind, replications are ideal to help validate the results and help detect any fluke results. 

The power of efficiency in a DOE is within hidden replication. However, the may be instances with a block design is incomplete when it isn't possible to apply all treatments in every block. 

More treatments takes more time and money but offers the most information. It's a trade-off between the amount of Type I and II error you can afford to risk along with time and money. 

The more factors and levels you have, the more combinations are possible and thus adding more time and money to the project, unless you choose to take more risk and reduce the runs by using a fractional design. Of course, being able to adjust setting of more variables and playing more with factors is great, but it comes with a price. 

The guide below shows that the amount of 'levels' to the power of the number of 'factors' is the number of combinations of treatments for a full factorial design. 

For example, with two factors (inputs) each taking two levels, a factorial DOE will have four combinations. With two levels and two factors the DOE is termed a 2×2 factorial design. A memory tactic.... Levels lie low and Factors fly high A DOE with 3 levels and 4 factors is a 3×4 factorial design with 81 (3 4 = 81) treatment combinati ons. It may not be practical or feasible to run a full factorial (all 81 combinations) so a fractional factorial design is done, where usually half of the combinations are omitted.

STEPS to conduct a DOE:

• Define the objective for the DOE
• Select the process variables (independent and dependent)
• Determine DOE design - which d epends on resources (time & money) and amount of Type I and II error you're willing to accept
• Execute the design (randomization where possible)
• Verify results (r eplicate the tests if practical to help verify results)
• Interpret the results

The next step is to implement IMPROVEMENTS (that’s the goal….implementing improvements that matter)

Here are some characteristics of factorial experiments in general:

• A Response is the output and is the dependent variable
• Response = sum of process mean + var iation about the mean
• Factors are independent variables
• Variation about the mean is sum of factors + interactions + unexplained residuals (or experimental error)

ANOVA is used to decompose the variation of the response to show the effect from each factor, interactions, and experimental error (or unexplained residual). Statistical software will help manage the entire DOE.

• Enter the factors
• Set the levels (at least two for each factor)
• Determine how many runs (full factorial, fractional factorial)
• Run the experiment at each treatment level
• Enter the response for each treatment level
• Use statistical software to use ANOVA on the data
• Continue to refine until prediction equation is obtained
• IMPROVE the KPIV's
• Last phase is CONTROL the KPIV's

Other methods of experimentation such as "trial and error" or "one factor at a time (OFAT)" are prone to waste, will provide less information and will not provide a prediction equation. These may seem easier to run and get results but the risk is a less robust solution and decisions made on a poor experiment. These input factors behave to create an output, the team needs to make improvements in the IMPROVE phase that control the inputs. Controlling the input factors will provide the desired response. The DOE will quantify the factor interactions and offer a prediction equation. ANOVA will help indicate which factors and combinations are statistically significant and which are not thus giving direction to the priority of improvements. DOE Assumptions since ANOVA is used to analyze the data:

• The residuals are independent
• The residuals have equal variance
• The residuals are normally distributed
• All inputs (factors) are independent of one another

Most prediction equations will be linear and reliable when using only two levels. This saves time and money while obtaining a prediction equation. Prediction equations are useful to analyze what-if scenarios. Many times data can not be collected at all levels and factors so a prediction equation can be used to estimate the output. The input factors are x's and the response is Y-hat.

Full Factorial DOE

The following are characteristics of a Full Factorial DOE:

• Usually results is large number of tests. If the number of parameters is large then the number of test becomes significantly large (there are more and more interaction combinations and possibilities).
• Testing every combination of factor levels.
• Captures all interactions which of course is nice to have but this comes at a cost and time.

For instance, if there 9 factors and 3 levels for each factor that the team wants to test, then that is 3 9 = 19,683 runs to determine all the interactions! 

3 * 3 Full Factorial DOE

Using the same vehicle throughout and maintaining all external variables as constant as possible a study is being created to find a prediction equation for the miles per gallon (MPG). There are 27 runs needed to bring out all the interactions (3 3 ). The team has determined that coefficient of friction of surface, ambient temperature, and tire pressure are three critical input factors (KPIV's) to study. The goal isn't always to maximize MPG but to understand the impact on vehicle MPG based on these factors. The problem statement may be to improve the accuracy of MPG claims on this specific vehicle.

The table below summarizes the three levels chosen for each of the three factors.

How many trials are required if you want to run a Full Fractional DOE with 5 factors at 4 levels each?

ANSWER: 4 5 = 1,024 trials (this could be impractical...thus look into the option below).

Fractional Factorial DOE

A Fractional Factorial experiment uses subset of combinations from a Full Factorial experiment. They are applicable when there are too many inputs to screen practically or cost or time would be excessive. 

This type of DOE involves less time than One-Factor at a Time (OFAT) and a Full Fractional Factorial but this choice will result in less data and some interactions will be confounded (or aliased). This means that the effect of the factor cannot be mathematically distinguished from the effect of another factor.  

Most processes are driven by main effects and lower order interactions so choose the higher order interactions for confounding. Lower confounding is found with higher resolution.

If a half fractional factorial experiment is determined to be most practical and economical where there are two levels and five factors then there will be a combination of 16 runs analyzed. Usually higher order interactions are omitted to focus on the main effects.

One-Way Experiment:  involves only one factor. 

Response (Y, KPOV): the process output linked to the customer CTQ. This is a dependent variable.

Factor (X, KPIV): uncontrolled or controlled variable whose influence is being studied. Also called independent variables. Inference Space: operating range of factors under study Factor Level: setting of a factor such as 1, -1, +, -, hi, low. Treatment Combination (run): setting of all factors to obtain a response Replicate: number of times a treatment combination is run (usually randomized). Replication is done to estimate the Pure Trial Error to the Experimental Error. Replication is very important to under confounding and interactions. ANOVA : Analysis of Variance Blocking Variable: Variable that the experimenter chooses to control but is not the treatment variable of interest. Interaction: occurrence when the effects of one treatment vary according to the levels of treatment of the other effect.

Main effect: estimate of the effect of a factor that is independent from any of the other factors.

Collinear:  variables that are linear combinations of one another. Two perfectly collinear variables with an exact linear relationship will have correlation of -1 or 1. Confounding: variables that are not being controlled by the experimenter but can have an effect on the output of the treatment combination being studied. It describes the mixing of estimates of the effects from the factors and interactions. Two (or more) variables are confounded if effects of two or more factor aren't separable.

Sensitivi ty:  refers to the ability to identify significant treatment differences in the response variable.

Covariate:  Factors that change in an experiment that were not planned to change .

Explaining DOE

This is a lengthy video but it slowly but clearly teaches the concepts and jargon and then jumps into an example at the end. The prelude to the example helps put all the pieces together before diving into an example.

Other Types of DOE's

Taguchi's design.

Taguchi's Design uses orthogonal arrays to estimate the main effects of many levels or even mixed levels. A selected and often limited group of combinations are investigated to estimate the main effects.

The goal is to find and develop a parameter that can improve a performance characteristic. It can be used to look for alternative materials or design methods that deliver equivalent or better performance.

The intent is to reduce the quality loss to society. Taguchi has the concept of loss function and assumes losses when a process doesn't meet a target value. The losses are from the variation of a process output. He states that losses rise quadratically as they move from the target value to the LSL/USL (may be one or the other or both).

Signal to Noise (S/N) ratios are used to improve the design. Ideally, the output from the design should not react to variation from the noise factors. 

Plackett-Burman

Plackett-Burman - two level fractional factorial design that analyzes only a few selected combinations to evaluate on main effects and no interactions.

Response Surface Methodology

Response Surface Methodology (RSM)  is used to study multiple factors although two are normally done.

RSM creates a map of the response from running a series of full factorial DOE's and comes up equations that describe how the factors affect the response.

RSM designs are used to refine processes after an experiment such as Plackett-Burman has identified the vital main effects. Then one can determine the settings of the factors to achieve of the desired response. 

Circumscribed Central Composite (CCC), Face Centered Composite (CCF), and Inscribed Central Composite (CCI) are designs that require 5 factor levels. As you can see in their names, these are all varieties of central composite designs. They are (and also Box-Behnken) RSM designs. 

RSM's may have a 3D response. Consider the following equation that came from an experiment:

y = 3.2 + 4.5x 1  + 5.2x 2  + 9x 2 2  + 8.2x 1 x 2

The formula contains two slope components (4.5x 1  and 5.2x 2 ), and curve component (9x 2 2 ) and a twist component (8.2x 1 x 2 )

Box-Behnken

Similar to the Face Centered Composite (CCF) in which it requires 3 factor levels

Considered a Response Surface Methodology Design

Circumscribed Central Composite  (CCC)

Similar to Inscribed Central Composite (CCI) design which is also a higher order design and both require 5 factor levels.

Alpha, α, = [2 k ] 1/4  where k = number of factors

Alpha is the distance of each axial point (star point) from the center in a central composite design. A value <1 puts the axial points in the cube; a value =1 puts them on the faces of the cube; and a value >1 puts them outside the cube.

Return to the IMPROVE phase Get access to all the material within this site Search active Six Sigma job openings Return to Six-Sigma-Material Home Page

Recent Articles

Process capability indices.

Oct 18, 21 09:32 AM

Six Sigma Calculator, Statistics Tables, and Six Sigma Templates

Sep 14, 21 09:19 AM

Six Sigma Templates, Statistics Tables, and Six Sigma Calculators

Aug 16, 21 01:25 PM

• Templates + Calcs
• Training Slides
• Six Sigma Jobs

Practice Exam

• Six Sigma Courses

FULL Site Access

• Subscription

A few Member Pages

• Project Mgmt

Project Pitfalls

• Project Acceleration
• Power/Sample Size
• Hyp. Test Flowcharts
• Cost of Inventory
• Financial Savings
• Icebreakers
• Multi-Vari Study
• Fishbone Diagram
• Normalized Yield
• Spearman's Rho
• DEFINE Phase
• MEASURE Phase
• ANALYZE Phase
• IMPROVE Phase
• CONTROL Phase
• LEAN Manufacturing

Basic Statistics

• Predictive Maint.

Six Sigma Careers

• MBB Training
• BLACK BELT Training
• GREEN BELT Training
• YELLOW BELT Training
• Certifications
• Stats Software
• Quality Gurus
• Six Sigma Books

Six Sigma, Six Sigma Training, Courses, Calculators, Certification

Aug 15, 21 10:27 PM

Site Membership LEARN MORE

Templates, Tables & Calculators

Six Sigma Slides

Green Belt Program (1,000+ Slides)

Cost of Quality

Process Mapping

Capability Studies

Cause & Effect Matrix

Multivariate Analysis

Central Limit Theorem

Confidence Intervals

Hypothesis Testing

1-Way ANOVA

Correlation

Control Plan

MTBF and MTTR

Error Proofing

Line Balancing

Yield Metrics

Sampling Methods

Data Classification

... and more

Need a Gantt Chart?

Copyright © 2024 Six-Sigma-Material.com. All Rights Reserved. | Privacy Policy

Design of Experiments Uncovers Hidden Factors in Production Processes

Image Source: gorodenkoff / iStock/Getty Images Plus

Design of Experiments helps manufacturers improve products and processes faster and more efficiently by testing multiple factors at once.

At its core, DOE tests how several factors work together to affect a product, instead of changing just one thing at a time. This gives a fuller picture of what influences the end result.

The benefits of DOE in manufacturing are numerous:

• Efficiency: By allowing multiple factors to be tested concurrently, DOE significantly reduces the number of experiments needed. This translates to savings in time and resources, which translates to faster product development and process optimization.
• Interaction Detection: DOE finds hidden connections between factors that simpler experimental designs can miss. Because multiple variables often influence a product’s final level of quality, the ability to spot complex interactions matters.
• Broad Applicability: From automotive to food production, DOE can be applied across various manufacturing sectors. It's equally effective for optimizing chemical formulations, machine settings or product compositions.
• Robustness: Beyond finding optimal settings, DOE helps identify process parameters that are less sensitive to uncontrollable variations. This leads to more stable manufacturing processes and consistent product quality.
• Predictive Power: After conducting a DOE, manufacturers can predict outcomes for factor combinations not directly tested — which is invaluable for further optimization without additional testing.

The DOE process typically follows these steps:

• Define the objective (e.g., reducing defects, improving strength)
• Identify factors to be varied
• Design the experiment using statistical software
• Conduct the experiments and collect data
• Analyze results to determine optimal settings and factor interactions

One of DOE's strengths is its efficiency. A well-designed experiment can provide comprehensive insights with a surprisingly small number of runs. For instance, a study involving 5-6 factors might require only 16-32 experimental runs, a fraction of what would be needed in exhaustive testing.

As manufacturing becomes increasingly complex with the advent of technologies like 3D printing and smart factories, DOE provides a structured approach to understanding and optimizing these complex systems.

Share This Story

Stay in the know with quality’s comprehensive coverage of the manufacturing and metrology industries., enewsletter | website | emagazine.

Copyright ©2024. All Rights Reserved BNP Media.

Design, CMS, Hosting & Web Development :: ePublishing

Researchers Have Discovered 2 New Dementia Risk Factors. Here's What They Are.

Wellness Reporter, HuffPost

Strides have been made in the world of dementia research compared with even just a few years ago. There’s now a blood test that can diagnose Alzheimer’s accurately 90% of the time , and more is understood about the factors (many of which are lifestyle habits) that can put you at higher risk for the condition.

In a new dementia report published in The Lancet journal by researchers who are part of The Lancet Commission, two new modifiable risk factors have been identified: high cholesterol after 40 and untreated vision loss.

In 2020, these same researchers determined 12 modifiable risk factors that are known to put folks at higher risk of developing dementia. These are:

• Physical inactivity
• Excessive alcohol consumption
• Air pollution
• Head injury
• Infrequent social contact
• Less education
• Hypertension
• Hearing impairment

According to the report, these 12 factors, along with the two new ones, account for 49% of dementia cases across the world. Researchers determined these two new risk factors by looking at recent meta-analyses and studies on the topics; they looked at 14 papers on vision loss and 27 on high cholesterol.

“It makes a lot of mechanistic sense,” said Dr. Arman Fesharaki-Zadeh , a behavioral neurologist and neuropsychiatrist at Yale Medicine in Connecticut. “A lot of these factors are very much interrelated.” (Fesharaki-Zadeh is not affiliated with the report.)

“There are many sources of vision loss, of course, but it tends to be a lot more common in folks who have metabolic risk factors such as high blood pressure, such as poorly controlled diabetes, such as high cholesterol, which is the other risk factor [identified in the report],” he said.

Moreover, vision is our primary sensory organ — it’s how we process the world around us — and when you can’t see clearly, you’re less likely to spend time doing brain-boosting activities like puzzles, reading or even spending time with other people, said Fesharaki-Zadeh. And these activities are known to help prevent dementia.

When it comes to high LDL cholesterol (the so-called bad cholesterol), it can lead to the hardening of the blood vessels in the heart and brain, Fesharaki-Zadeh said, adding that high blood pressure and uncontrolled diabetes also affect the blood vessels.

This can make it more difficult for oxygen to get to the brain, which over time can lead to neuron damage — “and dementia is essentially an end product of the neurons dying out, so it’s a neurodegenerative process,” Fesharaki-Zadeh explained.

“I can’t tell you how often I see in our patient populations, especially folks above the age of 60, there are certain parts of the brain that are more vulnerable to damage ... and these are the areas that are especially vulnerable to hardening of blood vessels. Someone who has ... high cholesterol, the correlation between that and hardening of blood vessels is quite high, and we see it in our clinical setting very frequently as well.”

“The saying that I like to use with patients quite often is what affects your heart will affect your brain, and we see that time and time again,” the doctor said.

You can lower your risk. First, have a good medical team and primary care doctor.

“I cannot highlight the importance of a collaborative model between primary care physicians and specialties,” said Fesharaki-Zadeh. Having a primary care doctor who understands your health and is willing to share pertinent information with specialists, like cardiologists and neurologists, will help you stay on top of any issues putting your well-being at stake.

Your primary care doctor should also be proactively working to help you control the risk factors — like high cholesterol and high blood pressure — whether that’s through medication, diet or exercise.

Fesharaki-Zadeh said you and your doctor should focus on these lifestyle changes as early as possible, at least in midlife, not when you’re at the point when dementia starts to show up.

“The front line of medical care are primary care physicians. These are the folks that, by having early discussions ... can go a long distance to prevent the onset of dementia,” he explained.

There are also tests that can detect early signs of neurodegeneration and genetic markers of the disease. A primary care doctor can help you learn about these options.

“Up to 40% of dementias are potentially preventable,” he added, but it’s worth noting that dementia can also be genetic, which makes prevention trickier. But someone who is diagnosed with dementia or mild cognitive impairment can benefit from managing these risk factors, too.

“The research is also showing that if you have two groups of individuals, someone who has comorbid metabolic diseases such as hypertension, high cholesterol, diabetes, versus somebody who doesn’t, and both of these individuals have dementia, the rates of progression of dementia in somebody who doesn’t have metabolic risk factors tend to be slower,” explained Fesharaki-Zadeh.

It’s never too late to make changes and corrections, he noted, whether you’re a young, seemingly healthy person, in your 80s or 90s, or someone who has already been diagnosed with dementia.

Our brains are highly malleable, Fesharaki-Zadeh said. So if you decide to make healthy lifestyle changes at any point, your brain will respond and be healthier for it.

From Our Partner

Huffpost shopping’s best finds, more in life.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 05 August 2024

Nucleophosmin 1 promotes mucosal immunity by supporting mitochondrial oxidative phosphorylation and ILC3 activity

• Rongchuan Zhao 1 , 2   na1 ,
• Jiao Yang   ORCID: orcid.org/0000-0002-1553-5425 3   na1 ,
• Yunjiao Zhai 4 ,
• Hong Zhang 1 ,
• Yuanshuai Zhou 1 , 2 ,
• Lei Hong 1 , 2 ,
• Detian Yuan 5 ,
• Ruilong Xia 6 ,
• Yanxiang Liu 3 ,
• Jinlin Pan 1 , 2 ,
• Shaheryar Shafi   ORCID: orcid.org/0000-0002-1936-4838 1 , 2 ,
• Guohua Shi 1 , 2 ,
• Ruobing Zhang 1 , 2 ,
• Dingsan Luo 1 ,
• Jinyun Yuan 1 ,
• Dejing Pan 7 ,
• Changgeng Peng   ORCID: orcid.org/0000-0002-6707-2717 6 , 8 ,
• Shiyang Li   ORCID: orcid.org/0000-0003-1487-4991 4 &
• Minxuan Sun   ORCID: orcid.org/0000-0003-2033-6268 1 , 2  

Nature Immunology ( 2024 ) Cite this article

1977 Accesses

23 Altmetric

Metrics details

• Innate lymphoid cells
• Tumour immunology
• Ulcerative colitis

Nucleophosmin 1 (NPM1) is commonly mutated in myelodysplastic syndrome (MDS) and acute myeloid leukemia. Concurrent inflammatory bowel diseases (IBD) and MDS are common, indicating a close relationship between IBD and MDS. Here we examined the function of NPM1 in IBD and colitis-associated colorectal cancer (CAC). NPM1 expression was reduced in patients with IBD. Npm1 +/− mice were more susceptible to acute colitis and experimentally induced CAC than littermate controls. Npm1 deficiency impaired the function of interleukin-22 (IL-22)-producing group three innate lymphoid cells (ILC3s). Mice lacking Npm1 in ILC3s exhibited decreased IL-22 production and accelerated development of colitis. NPM1 was important for mitochondrial biogenesis and metabolism by oxidative phosphorylation in ILC3s. Further experiments revealed that NPM1 cooperates with p65 to promote mitochondrial transcription factor A (TFAM) transcription in ILC3s. Overexpression of Npm1 in mice enhanced ILC3 function and reduced the severity of dextran sulfate sodium-induced colitis. Thus, our findings indicate that NPM1 in ILC3s protects against IBD by regulating mitochondrial metabolism through a p65-TFAM axis.

Similar content being viewed by others

Intestinal epithelium-derived BATF3 promotes colitis-associated colon cancer through facilitating CXCL5-mediated neutrophils recruitment

PD-1 signaling facilitates activation of lymphoid tissue inducer cells by restraining fatty acid oxidation

OLFM4 deficiency delays the progression of colitis to colorectal cancer by abrogating PMN-MDSCs recruitment

Inflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are characterized as chronic and recurring ailments of the gastrointestinal tract 1 , which is considered a high risk of colitis-associated colorectal cancer (CAC) 2 , 3 , 4 . The precise pathogenesis of IBD remains unknown, but hypotheses include immune response disorders, alterations in intestinal microbiota, genetic susceptibility and environmental factors 2 , 5 .

Myelodysplastic syndrome (MDS) is a hematopoietic stem cell disorder characterized by deficient hematopoiesis, cytopenia of peripheral blood and a predisposition to acute myeloid leukemia (AML) 6 , 7 , 8 . The cause of MDS is linked to the presence of acquired chromosomal abnormalities and genetic mutations that alter oncogene and tumor suppressor gene function 9 . Since the first report of seven patients with both IBD and MDS in 1997, numerous cases of concurrent IBD and MDS have been documented 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 . Case studies of patients with IBD indicate a high incidence of AML/MDS in patients with IBD 16 . A high prevalence of IBD was also found in a large cohort of patients with MDS, suggesting a close association between IBD and MDS 15 .

Mutations in nucleophosmin 1 (NPM1, also known as B23, numatrin 1 or NO38) are associated with a high risk of MDS and AML 6 , 17 . NPM1 was identified as a nucleolar phosphoprotein with multiple functions and binding partners 18 . NPM1 interacts with many partners in distinct cellular compartments, including nucleolar factors, transcription factors and histones. NPM1 is the most frequently mutated gene in patients with AML 19 , 20 , accounting for ~60% of patients with a normal karyotype and 35% of total cases 21 , 22 . However, whether NPM1 regulates IBDs remains unknown.

Innate lymphoid cells (ILCs) were characterized as a family of heterogeneous lymphocytes that originate from common lymphoid progenitors in the bone marrow but with the absence of variable antigen receptors 23 . Group three ILCs (ILC3s) are the most abundant subgroup of ILCs in the gut and are the primary source of interleukin-22 (IL-22). ILC3s expressing the transcription factors retinoid-related orphan receptor gamma t (RORγt) 24 and aryl hydrocarbon receptor 25 , 26 produce IL-22, which triggers the synthesis of antimicrobial peptides, such as RegIIIβ and RegIIIγ, by epithelial cells 27 , 28 . Thus, ILC3s are at the beginning of a pathway that promotes immunity to infection. In a colon cancer model, Il22 −/− mice were observed to undergo accelerated tumorigenesis compared to wild-type (WT) mice 29 , suggesting a potential protective role for ILC3s in gut homeostasis.

In this study, we investigate the protective role of NPM1 in gut homeostasis and in the prevention of colitis. Using Npm1 -haploinsufficient ( Npm1 +/− ) mice, we observed increased susceptibility to colitis and colitis-associated colorectal cancer. NPM1 was abundant in ILC3s and was essential for IL-22 production in response to dextran sulfate sodium (DSS)-induced colitis. Conditional deletion of Npm1 in the ILC3 lineage exacerbated colitis and decreased protective IL-22 secretion. Additionally, heterozygous deletion of Npm1 in ILC3 dysregulated mitochondrial homeostasis, including decreased mitochondrial biogenesis and oxidative phosphorylation (OXPHOS). Mechanistically, we found that NPM1 acted as a transcription cofactor that bound p65 and stimulated mitochondrial transcription factor A ( Tfam ) transcription in DSS-induced colitis. Thus, our findings demonstrated that NPM1 regulates mitochondrial function and IL-22 production in ILC3s through the p65-TFAM axis, promoting gut homeostasis and protection against IBD.

NPM1 deficiency leads to increased susceptibility to colitis

Patients with UC exhibited a decreased abundance of NPM1 in the colon compared to controls (Fig. 1a and Supplementary Table 1 ). We also observed a trend in reduced NPM1 in patients with CD; however, the reduction compared to controls was not significant (Fig. 1a ). Further single-cell RNA-sequencing (scRNA-seq) analysis ( GSE182270 ) on colonic biopsies of patients with UC and healthy control (HC) 30 indicated that the expression of NPM1 decreased mainly in ILC3s, macrophages, natural killer T cells (NKT), cytotoxic T cells, regulatory T cells (T reg ) and Paneth cells in patients with UC (Extended Data Fig. 1a,b ). NPM1 mRNA abundance was also significantly reduced in patients with high-grade colon adenocarcinoma (COAD; stages III and IV), compared to those with low-grade COAD (stages I and II; Extended Data Fig. 1c ). Analysis of The Cancer Genome Atlas database revealed that lower NPM1 mRNA correlated with worse overall survival in patients with either COAD or rectum adenocarcinoma (READ; Extended Data Fig. 1d,e ). These findings suggested that NPM1 may be involved in the pathology of IBD, especially UC, and may contribute to tumorigenesis.

a , Immunohistochemistry of NPM1 in colon tissue from patients with IBD (UC, n  = 22 individual patients and CD, n  = 20 individual patients) and non-IBD ( n  = 29 individual patients) controls. Scale bars = 10 μm. Immunohistochemistry score of NPM1. Statistical differences were determined by the Mann–Whitney test (** P  < 0.01). b – f , Npm1 +/− and control Npm1 + / + mice were administered 2.5% DSS for 7 days, followed by 3 days of recovery (H 2 O). Body weight ( b ), DAI (a score of inflammation) in the colon ( c ), colon length on day 10 ( d , e ) and colon histopathology on day 10 ( f ) were analyzed ( n  = 5 individual mice). Scale bars = 500 μm (left) and 100 μm (right). ( g , h ) RT–PCR analysis of mRNA abundance of Reg3b and Reg3g ( g ) and S100a8 and S100a9 ( h ) in the whole colon of mice at day 5 of administration of 2.5% DSS ( n  = 4 individual mice). i , Diagram of AOM/DSS CAC model. j , Total number of tumors and number of tumors larger than 2 mm in Npm1 + / + and Npm1 +/− mice ( n  = 5 individual mice). k , Representative images of colons with tumors from Npm1 + / + and Npm1 +/− mice on day 65 of the AOM/DSS CAC model. l , Histopathology of representative colon tumors from Npm1 + / + and Npm1 +/− mice on day 65 of the AOM/DSS CAC model. Scale bars = 100 μm. Data in e , g , h and j are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t test (** P  < 0.01, *** P  < 0.001 and **** P  < 0.0001), unless otherwise indicated. Ctrl, control; NS, not significant; CAC, colitis-associated colorectal cancer.

Source data

To explore the putative contribution of NPM1 in gastrointestinal homeostasis and inflammation, we generated Npm1- haploinsufficient ( Npm1 +/− ) mice (Supplementary Fig. 1a–c ) and confirmed reduced abundance of NPM1 in the colon (Extended Data Fig. 1f–h ). Note that homozygous knockout was lethal. In the absence of injury, colon length and histology were similar between WT mice and Npm1 +/− mice (Extended Data Fig. 1i,j ). Concurrently, the organogenesis of secondary lymphoid structures, including Peyer’s patches (PP) and mesenteric lymph nodes (MLN), as well as solitary intestinal lymphoid tissue, was unaffected by Npm1 haploinsufficiency (Extended Data Fig. 1k–n ). Given the critical role of NPM1 in MDS and AML, we also examined the change of bone marrow (BM) cells in Npm1 +/− mice (Supplementary Fig. 2a ). Results indicated that the ratio of Lin − c-Kit + (LK) cells, Lin − Sca1 + c-Kit + (LSK) cells and LS low K low cells in the BM of Npm1 +/− mice was elevated compared with that of Npm1 + / + mice both in steady state and DSS-induced colitis conditions (Extended Data Fig. 1o,p ), which is a characteristic phenotype of MDS 21 , 31 , 32 . Meanwhile, within the LK cell population, the proportion of granulocyte–macrophage progenitors (GMP) increased in Npm1 +/− mice, especially in a steady state (Extended Data Fig. 1o,p ). Using a DSS colonic injury model (2.5% wt/vol for 7 days), we found that Npm1 +/− mice had greater body weight loss and a greater increase in disease activity index (DAI), a marker of inflammation, compared to littermate controls (Fig. 1b,c ). On day 10 (3 days into the recovery period), NPM1 deficiency exacerbated inflammation as indicated by reduced colon length and increased epithelial injury, submucosal edema and leukocyte infiltration in the colon (Fig. 1d–f ). Additionally, at day 5 of DSS exposure, expression of genes encoding antimicrobial peptides ( Reg3b and Reg3g ) was reduced and calprotectin ( S100a8 and S100a9 ), a marker of inflammation, was altered in colons of Npm1 +/− mice (Fig. 1g,h ). We also established a trinitrobenzene sulfonic acid (TNBS)-induced colitis model and evaluated the progress of colitis in WT and Npm1 +/− mice. As anticipated, Npm1 +/− mice also exhibited reduced colon length and enhanced inflammation, together with greater body weight loss and increased DAI (Extended Data Fig. 1q–u ). Collectively, these data indicated that NPM1 has a protective role in the mouse colitis model.

NPM1 inhibits colitis-associated colon tumorigenesis

Patients with IBD have a high risk of developing CAC 33 , 34 . To investigate the role of NPM1 in CAC development, we subjected WT mice and Npm1 +/− mice to an azoxymethane (AOM)/DSS colon tumor model (Fig. 1i ). By the end of the third cycle of DSS treatment, Npm1 +/− mice failed to recover body weight and exhibited increased DAI (Extended Data Fig. 1v,w ). Compared to WT mice, Npm1 +/− mice developed more tumors and a greater number of larger tumors, indicative of a higher tumor burden (Fig. 1j–l ). In addition to CAC, which is preceded by chronic inflammation, sporadic colorectal cancer (CRC) is a form of CRC that is often caused by mutations in the gene APC 35 . To examine the role of NPM1 in sporadic CRC, we crossed Npm1 +/− mice with Apc min / + mice and fed them a Western diet to accelerate tumorigenesis. Results showed that there were no significant differences in colonic tumor load between Apc min / + mice and Npm1 +/− Apc min / + mice (Extended Data Fig. 1x ), suggesting that sporadic CRC arising from APC mutations does not involve NPM1. Collectively, these findings indicated that NPM1 has a pivotal role in impeding colitis-associated colon tumorigenesis by restricting tumor development and growth.

Protection against colitis requires NPM1 in hematopoietic cells

Given that Npm1 is expressed by many types of cells and decreased under pathological conditions (Extended Data Figs. 1b and 2a ), it is unclear whether the exacerbated colitis in Npm1 +/− mice is due to defects in hematopoietic or nonhematopoietic cells, particularly colonic epithelial cells. Thus, we established BM chimeras with Npm1 deficiency in these distinct cellular populations (Extended Data Fig. 2b ). After a 7-day DSS treatment, mice receiving Npm1 -haploinsufficient BM exhibited more severe colitis compared to mice receiving Npm1 WT BM cells. However, when the same donor BM was used regardless of the genotype of the host mice, there were no significant differences in body weight, colon length or histological features (Extended Data Fig. 2c–f ), suggesting that the hematopoietic compartment is the main functional compartment for NPM1. We also detected the expression of tight junction genes (including Tjp1 , Tjp2 , Cldn2 and Cldn3 ) in epithelial cells, which are pivotal for the maintenance of intestinal barrier function 36 . With the exception of Cldn3 , which is diminished in Npm1 -haploinsufficient mice under physiological conditions, the expression of other tight junction genes remains relatively unchanged between two groups of mice in both physiological and pathological conditions (Extended Data Fig. 2g–j ). We also generated Npm1 flox / flox mice (Supplementary Fig. 1d–f ) and crossed them with Villin cre / + mice to directly assess a role in protection against colitis for NPM1 in colonic epithelial cells. However, there was no obvious alteration in colon length and histological features between Villin cre / + Npm1 flox / flox mice and control mice (Extended Data Fig. 2k–m ). Taken together, these data showed that impaired gut homeostasis and exacerbated inflammation in Npm1 +/− mice are mainly caused by the heterozygous deletion of Npm1 in the hematopoietic compartment.

NPM1 is critical for maintaining IL-22-producing ILC3s

Subsequently, we investigated the type of gut immune cells involved in limiting gut inflammation by NPM1. The ratio of macrophages, neutrophils, eosinophils and dendritic cells (DCs) infiltrated in intestinal lamina propria leukocytes (LPLs) exhibited few changes between WT and Npm1 +/− mice in steady state (Supplementary Fig. 2b and Extended Data Fig. 3a–d ). However, in DSS-induced colitis, an elevation of these cells was observed in Npm1 +/− mice compared to WT mice (Extended Data Fig. 3e–h ). It’s known that infiltration of myeloid cells into the intestinal lamina propria is considered a common cause of progressive colitis 37 . Furthermore, clearance of CD11b + myeloid cells failed to rescue the exacerbated enteritis in Npm1 +/− mice, suggesting that NPM1 in myeloid cells was insufficient to regulate intestinal inflammation (Extended Data Fig. 3i–m ). Likewise, evaluation of T cells (T H 17, T reg and γδT cells) coupled with comparable colitis in two genotype mice after deletion of CD3 + T cells indicated that exacerbated colitis in Npm1- haploinsufficient mice was not attributed to T cells (Supplementary Fig. 2c and Extended Data Fig. 3n–x ).

We then investigated the effect of Npm1 haploinsufficiency on colonic ILC3s (Supplementary Fig. 2d ). The population of colonic ILC3s and IL-22 + ILC3s decreased in Npm1 +/− mice compared to Npm1 + / + mice after DSS administration, suggesting that haploinsufficient of Npm1 affects ILC3 expansion and function (Fig. 2a–e ). Additionally, Npm1 +/− ILC3 exhibited similar alterations in TNBS-induced colitis (Extended Data Fig. 4a–d ). However, these changes were not observed under physiological conditions (Extended Data Fig. 4e–h ). Further analysis revealed that there were no evident alterations in proportions of NCR + ILC3 and CCR6 + ILC3 between WT and Npm1 +/− mice under physiological or pathological conditions (Extended Data Fig. 4i,j ). Moreover, isolated ILC3s from Npm1 +/− mice produced less IL-22 compared with ILC3s from WT mice after DSS administration (Fig. 2f ). In addition, the expression of Il22 was also decreased in isolated ILC3s from Npm1 +/− mice exposed to DSS, but the expression of Il22 was similar in both genotypes under steady state (Fig. 2g ). The decreased production of IL-22 in ILC3s may contribute to the observed dysregulation of Reg3b and Reg3g in Npm1 +/− mice in DSS-induced colitis (Fig. 1g ), and thus impaired intestinal microbiota homeostasis. There was a rapid decrease in observed operational taxonomic unit, Chao1 index and Shannon index in Npm1 +/− mice (Extended Data Fig. 4k–m ), indicating that microbiota diversity was repressed by Npm1 heterozygote deletion. Moreover, feces from Npm1 +/− mice and WT mice showed a remarkable change in bacterial composition (Extended Data Fig. 4n–q ). However, cohousing littermate Npm1 +/− mice still exhibited more pronounced exacerbation of enteritis compared to WT mice (Extended Data Fig. 4r–u ), indicating that changes in the gut microbiota are not the priori drivers of the exacerbated inflammation in Npm1 +/− mice but may instead contribute to a certain extent to the exacerbation of enteritis. Collectively, our results indicated that NPM1 is important for the protective function of ILC3s in the gut immune microenvironment.

a , Colon LPLs were isolated from Npm1 + / + and Npm1 +/− mice at day 5 of administration of 2.5% DSS. Analysis of ILC3s (live CD45 + Lin − RORγt + cells) and IL-22-producing ILC3s (live CD45 + Lin − RORγt + IL-22 + cells) by flow cytometry. Numbers indicate percentages of cells in each outlined region. b , c , The proportion of CD45 + cells that are ILC3s ( b ; n  = 6 individual mice) and the proportion of IL-22 + ILC3s in the total ILC3 population ( c ; n  = 5 individual mice) in LPLs of Npm1 + / + and Npm1 +/− mice after DSS administration are shown. d , e , Number of ILC3s ( d ) and IL-22 + ILC3s ( e ) in LPLs of Npm1 + / + and Npm1 +/− mice after DSS administration are depicted ( n  = 5 individual mice). f , ILC3s, isolated by cell sorting from LPLs of Npm1 + / + and Npm1 +/− mice after DSS administration, were analyzed by ELISA for IL-22 ( n  = 3 individual mice). g , Relative mRNA abundance of Il22 in ILC3s, isolated by cell sorting from the LPL of Npm1 + / + and Npm1 +/− mice exposed to 2.5% DSS or water (steady state), was analyzed. The results are shown relative to the amount in cells from Npm1 + / + mice exposed to water (steady state; n  = 6 individual mice). h – l , Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice were administered 2.5% DSS for 7 d followed by 3 d of recovery. Body weight ( h ), DAI ( i ), colon length ( j , k ) and colon histopathology on day 10 ( l ) were analyzed ( n  = 5 individual mice). Scale bars = 500 μm (left) and 100 μm (right). m , n , LPLs were isolated from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice after 5 days of administration of 2.5% DSS ( n  = 5 individual mice). The proportion of CD45 + cells that are ILC3s ( m ) and the proportion of IL-22 + ILC3s in the total ILC3 population ( n ) are shown. o , p , The number of ILC3s ( o ) and IL-22 + ILC3s ( p ) in LPLs of Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice after DSS administration are depicted ( n  = 5 individual mice). q , Relative mRNA abundance of Il22 in ILC3s, isolated by cell sorting from the LPL of Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice at day 5 of administration of 2.5% DSS was analyzed ( n  = 4 individual mice). r , ILC3s, isolated by cell sorting from LPLs of Npm1 flox / flox and Rorc cre / + Npm1 flox / flox , were analyzed by ELISA for IL-22 ( n  = 3 individual mice). s , t , IL-22 production by MNK3 cells after stimulation with IL-1β and IL-23 in vitro by flow cytometry ( s ; n  = 5 individual mice) and by ELISA ( t ; n  = 3 individual mice). Data in b – g , k and m – t are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t test (* P  < 0.05, ** P  < 0.01, *** P  < 0.001 and **** P  < 0.0001). LPLs, lamina propria leukocytes.

To specifically decipher the cell-intrinsic role of Npm1 in colonic ILC3s, we generated Rorc cre / + Npm1 flox / flox mice that lack Npm1 on ILC3s and subjected the mice to DSS-induced colitis. The development of PP and MLN was unimpaired in Rorc cre / + Npm1 flox / flox mice (Extended Data Fig. 5a–c ). Frequencies of intestinal ILC3 and IL-22 + ILC3 in Rorc cre / + Npm1 flox / flox mice were also comparable with those of the control group (Extended Data Fig. 5d–g ). However, compared to Npm1 flox / flox mice, Rorc cre / + Npm1 flox / flox mice exhibited greater loss of body weight and increased DAI (Fig. 2h–i ), indicating exacerbated inflammation following DSS administration. When killed on day 10 (3 days after recovery), Rorc cre / + Npm1 flox / flox mice exhibited decreased colon length and greater features of colon injury (Fig. 2j–l ). The frequencies of colonic ILC3s and IL-22 + ILC3s were also decreased in Rorc cre / + Npm1 flox / flox mice after DSS administration (Fig. 2m–p ). Without development defects, heterozygous deletion of Npm1 in ILC3 also contributed to exacerbated enteritis and reduction of ILC3, which appears to be in a dose-dependent manner (Extended Data Fig. 5h–q ). Furthermore, the percentage of apoptotic ILC3s was increased in Rorc cre / + Npm1 flox / flox mice in DSS-induced colitis (Supplementary Fig. 2e and Extended Data Fig. 5r,s ). The proportion of CCR6 + ILC3 in total ILC3s was higher in Rorc cre / + Npm1 flox / flox mice than that in Npm1 flox / flox mice under pathological conditions, which was opposite in steady state (Extended Data Fig. 5t,u ). The proportion of interferon-γ (IFNγ)-producing ex-ILC3 was also unchanged between these two groups of mice with or without DSS administration (Supplementary Fig. 2f and Extended Data Fig. 5v–y ). Additionally, consistent with changes observed in Npm1 +/− mice, the increased infiltration of myeloid cells also existed in Rorc cre / + Npm1 flox / flox mice under pathological conditions (Extended Data Fig. 6a–h ). Because RORc-Cre will also delete Npm1 in conventional T cells and γδT cells, we examined the function of various T cell subsets and excluded their contributions to exacerbated colitis in Rorc cre / + Npm1 flox / flox mice by depleting T cells using a CD3 antibody (Extended Data Fig. 6i–q ). Moreover, isolated ILC3s from Rorc cre / + Npm1 flox / flox mice showed less Il22 expression and IL-22 production compared with Npm1 flox / flox ILC3s (Fig. 2q,r ). We also confirmed our findings in vitro using the ILC3 cell line, MNK3. MNK3 cells retain phenotypic and functional features characteristic of mouse primary ILC3s, including the production of IL-17A and IL-22 when stimulated with IL-23 and IL-1β 38 , 39 . Knockdown of Npm1 in MNK3 significantly suppressed the secretion of IL-22 upon stimulation (Fig. 2s,t ). Furthermore, Rorc cre / + Npm1 flox / flox mice developed more tumors compared with Npm1 flox / flox mice when subjected to AOM/DSS (Extended Data Fig. 6r–t ). Collectively, these data supported that NPM1 in ILC3s is critical for gut homeostasis under injury conditions and limiting inflammation.

NPM1 promotes mitochondrial gene expression in ILC3s

To uncover mechanisms by which NPM1 regulates ILC3 expansion and function, we performed RNA-seq (smart-seq2) of Live + Lin − CD45 low CD90 high LPLs 27 , 40 from colon of WT and Npm1 +/− mice with colitis induced by DSS treatment (Fig. 3a ). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed that the OXPHOS pathway is a top differentially modulated pathway in Npm1 -haploinsufficient mice compared to WT mice (Fig. 3b ). We also observed decreased expression of several genes encoding mitochondrial complex subunits of OXPHOS in Npm1 +/− mice, specifically those for electron transport chain (ETC) complex I ( mt-Nd3 , mt-Nd4 , mt-Nd4l and Ndufa12-ps ), complex IV ( mt-Co2 and mt-Co3 ) and complex V ( mt-Atp6 and mt-Atp8 ; Fig. 3c ) and confirmed these findings by RT–PCR (Fig. 3d ). However, the universal decrease of mtDNAs was not observed in epithelial cells, macrophages or T cells of Npm1 +/− mice (Extended Data Fig. 7a–c ). The abundance of NDUFB8 (complex I), mt-CO1 (complex IV) and mt-ATP6 (complex V) was also notably reduced due to Npm1 -knockdown in MNK3 cells (Fig. 3e ). These results indicated that NPM1 has a role in regulating the OXPHOS pathway in ILC3s.

a , RNA-seq analysis of colonic ILC3 isolated from Npm1 + / + and Npm1 +/− mice at day 5 of administration of 2.5% DSS ( n  = 3). b , KEGG pathway enrichment analysis of downregulated genes in Npm1 +/− mice ( n  = 3 individual mice). c , Heatmap of selected DEGs encoding proteins involved in OXPHOS in ILC3s between Npm1 + / + and Npm1 +/− mice ( n  = 3 individual mice). d , RT–PCR analysis of mRNA abundance of the indicated genes in ILC3s, isolated by cell sorting from the LPL of Npm1 + / + and Npm1 +/− mice at day 5 of administration of 2.5% DSS ( n  = 3 individual mice). e , Western blot showing the abundance of selected mitochondrial complex components in MNK3 cells. The samples were derived from the same experiment, and the blots were processed in parallel. f , Single-cell analysis of colonic samples of patients with UC from the GEO database ( GSE182270 ). Representative DEGs ( x axis) by cluster ( y axis) with dot size representing the fraction of cells within the cluster that express each gene and colors indicating the z -scaled expression of genes in cells within each cluster. g , KEGG pathway enrichment analysis of upregulated genes in NPM1 high ILC3s compared to NPM1 low ILC3s from the data of patients with UC. Data in d is representative of two independent experiments, shown as the mean ± s.e.m., and statistical significance was determined by two-way ANOVA (* P  < 0.05 and ** P  < 0.01).

To validate that NPM1 regulates OXPHOS in ILC3s in humans as well, scRNA-seq data of human colonic biopsies ( GSE182270 ) 30 was analyzed, and the ILC3 cluster was identified based on higher expression of KIT , RORC but lower expression of CTLA4, CD3D and CD3G (markers of T cells; Fig. 3f ). Although NPM1 is broadly expressed across all clusters, ILC3s were among those with comparatively high expression (Fig. 3f ). Similar to our mouse data, KEGG pathway enrichment analysis of differentially expressed genes (DEGs) between NPM1 high and NPM1 low ILC3s in patients with UC identified the OXPHOS pathway among the top five pathways regulated by NPM1 (Fig. 3g ). These findings suggested that altered cellular metabolism through the OXPHOS pathway in ILC3s represents a potential mechanism by which NPM1 activity influences UC.

Lack of NPM1 impairs mito-OXPHOS and biogenesis in ILC3s

According to the abovementioned results, mitochondrial OXPHOS is probably impaired in Npm1 -haploinsufficient ILC3s (Fig. 3c–e ). Therefore, we evaluated OXPHOS in isolated ILC3s from DSS-induced Npm1 +/− mice and WT mice. Heterozygous deletion of Npm1 in ILC3s reduced oxygen consumption rate (OCR) in response to DSS (Fig. 4a ). Compared to WT ILC3s, Npm1 +/− ILC3s exhibited a marked reduction in basal OCR, ATP production and maximal respiration (Fig. 4b–d ), indicating that mitochondrial OXPHOS in ILC3s was impaired by insufficient NPM1. However, such impaired mitochondrial function in Npm1 +/− ILC3s was not observed under physiological conditions (Extended Data Fig. 7d–g ). In the DSS model, mouse intestinal ILC3s exhibited a dramatic mitochondrial activation in the acute tissue damage phase (day 5) and then partially restored to a normal state in the repair phase (day 10; Extended Data Fig. 7h–k ). The inadequate mitochondrial activation of ILC3 in the acute phase caused by heterozygous deletion of Npm1 could lead to exacerbated colitis (Fig. 4b–d ). Besides, epithelial cells, macrophages and T cells in Npm1 +/− mice exhibited few differences in OXPHOS compared to those in WT mice in both steady state and DSS-induced colitis conditions (Extended Data Fig. 7l–t ). Moreover, the mitochondrial membrane potential of Npm1 +/− ILC3s was also reduced significantly compared with that of WT ILC3s only under pathological conditions (Fig. 4e,f and Extended Data Fig. 7u,v ). These results showed the importance of NPM1 in maintaining mitochondrial OXPHOS in ILC3s.

a – d , Cell mito stress test was performed with isolated colonic ILC3s from Npm1 + / + ( n  = 4 individual mice) and Npm1 +/− ( n  = 5 individual mice) mice. Representative OCR profile ( a ), basal OCR ( b ), ATP production ( c ) and maximal respiration ( d ) are shown. e , f , Mitochondrial membrane potential was assessed with the indicator 5,5ʹ,6,6ʹ-tetrachloro-1,1ʹ,3,3ʹ-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) ( e ) and TMRE ( f ) in isolated colonic ILC3s from Npm1 + / + and Npm1 +/− mice under DSS ( n  = 5 individual mice). g , Ultrastructural analysis of mitochondria by SEM of isolated colonic ILC3s from Npm1 + / + and Npm1 +/− mice. Scale bar = 1 μm. h , The number of mitochondria per cell was counted in SEM images ( n  = 10 fields per group). i , TOMM20 in ILC3s from Npm1 + / + and Npm1 +/− mice was detected by immunofluorescence staining. Scale bar = 5 μm. j – n , Npm1 + / + and littermate control Npm1 +/− mice were treated with bezafibrate (i.g., oral gavage) and administered 2.5% DSS for 7 days followed by 3 days of recovery. Body weight ( j ), DAI ( k ), representative colon images ( l ), colon length ( m ) and colon histopathology ( n ) are shown ( n  = 5 individual mice). Scale bars = 500 μm (up) and 100 μm (down). o , p , The proportion of CD45 + cells that are ILC3s ( o ) and the proportion of IL-22 + ILC3s in the total ILC3 population ( p ) in LPLs from mice of the indicated genotypes with or without bezafibrate treatment ( n  = 5 individual mice). q , r , Analysis of IL-22 production by isolated colonic ILC3s from mice treated with or without bezafibrate (10 mg kg −1 , i.g.; q ) and MNK3 ( r ) cells with or without bezafibrate (200 μM) addition by ELISA ( n  = 3 individual mice). Data in b – f , h , m , o – r are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t test (* P  < 0.05, ** P  < 0.01, *** P  < 0.001 and **** P  < 0.0001). BZ, bezafibrate.

To determine whether reduced transcription of OXPHOS genes in Npm1 -haploinsufficient ILC3s (Fig. 3c–e ) is associated with decreased mitochondrial biogenesis, we quantified the number of mitochondria using scanning electron microscopy (SEM; Fig. 4g ). We found fewer mitochondria per cell in Npm1 +/− ILC3s (Fig. 4h ). ILC3s from Npm1 +/− mice also had reduced staining of TOMM20, a mitochondrial protein (Fig. 4i ). Collectively, these data indicated that NPM1 has a critical function in maintaining mitochondria numbers and mitochondrial metabolism in ILC3s and that impairment of such metabolism represents a mechanism by which heterozygous deletion of Npm1 exacerbates DSS-induced colitis.

To confirm that mitochondrial biogenesis and function were impaired by Npm1 heterozygous deletion and that such impairment contributed to colitis severity, we used bezafibrate, an agonist of the transcription factors peroxisome proliferator-activated receptor (PPAR) γ coactivator 1α (PGC1α) that stimulates mitochondrial OXPHOS and biogenesis 41 , 42 . Compared with Npm1 +/− mice without bezafibrate treatment during the course of DSS administration, Npm1 +/− mice receiving bezabrifate exhibited greater recovery of body weight, greater reduction in DAI, longer colons and reduced inflammation (Fig. 4j–n ), suggesting that maintaining mitochondrial function through bezafibrate limited colitis severity in Npm1 +/− mice. However, bezafibrate had minimal impact on mice with sufficient NPM1 function, suggesting that both NPM1 and bezafibrate maintain mitochondrial function to limit colitis (Fig. 4j–n ). Similarly, bezafibrate succeeded in reversing the colitis in Rorc cre / + Npm1 flox / flox mice (Extended Data Fig. 7w–y ). Moreover, in Npm1 +/− mice exposed to DSS, bezafibrate resulted in increased percentages of total colonic ILC3s, IL-22 + ILC3s and IL-22 production by ILC3s (Fig. 4o–q ), suggesting that sufficient mitochondrial OXPHOS and biogenesis are required for ILC3 activity in DSS-induced colitis.

MNK3 also exhibited mitochondrial activation after IL-1β/IL-23 stimulation, which was regulated by NPM1 (Extended Data Fig. 8a–d ). Knockdown of Npm1 in MNK3 suppressed the secretion of IL-22 in response to IL-23 and IL-1β (Fig. 4r ). However, bezafibrate rescued ILC3 function in terms of IL-22 secretion in Npm1 -knockdown cells (Fig. 4r and Extended Data Fig. 8e,f ). In contrast, OXPHOS inhibitors oligomycin and rotenone suppressed the activation of MNK3 (Extended Data Fig. 8g,h ). However, the difference in Il22 expression between shNC and sh Npm1 MNK3 after OXPHOS inhibitor administration indicated that NPM1 may participate in other biological processes to sustain ILC3 activation (Extended Data Fig. 8g,h ). The tricarboxylic acid (TCA) cycle, a crucial component of mitochondrial metabolism, is known to participate in the activation of immune cells 43 . Because succinate is a substrate for the TCA cycle, its addition partially rescued the impaired ILC3 activation resulting from Npm1 heterozygous deletion (Extended Data Fig. 8i ). These results revealed that the defect in mitochondrial function resulting from Npm1 deficiency accounts for the impairment of ILC3 activation and function, leading to exacerbated colitis.

NPM1 regulates TFAM transcription by binding to p65

To uncover the molecular mechanism by which NPM1 regulates mitochondrial homeostasis of ILC3s, we immunoprecipitated MNK3 cells with or without stimulation (Fig. 5a ). Proteins associated with NPM1 were separated by SDS–PAGE and then silver stained. A band of ~70 kDa was enriched in stimulated MNK3 cells compared to unstimulated cells (Fig. 5b ). Mass spectrometry (MS) revealed that p65, a component of the nuclear factor kappa B (NF-κB) transcription factor, is the top candidate for ~70 kDa protein that co-immunoprecipitated with NPM1 (Fig. 5c ), which is consistent with a previous study reporting an interaction between NPM1 and p65 (also known as RelA), RelB and p50 (ref. 44 ). The two proteins, NPM1 and p65, were co-immunoprecipitated from stimulated MNK3 cells, confirming the MS findings and suggesting that the proteins interacted (Fig. 5d ). Immunofluorescence analysis of ILC3s revealed that p65 was localized in the cytoplasm and NPM1 was localized in the nucleus in the noninflammatory steady state, whereas p65 accumulated in the nucleus after DSS-induced colitis and colocalized with NPM1 (Fig. 5e ). Stimulation of MNK3 cells with IL-1β and IL-23 also promoted the accumulation of p65 in the nucleus (Fig. 5f ). These results indicated that inflammatory stimulation induces subcellular translocation of p65 and promotes the interaction between p65 and NPM1 in the nucleus of ILC3s.

a , Schematic diagram of protein–protein interaction analysis with NPM1. b , Silver-stained gel showing proteins that were immunoprecipitated with NPM1 and exhibited higher intensity in stimulated than unstimulated cells. Red rectangle shows bands that were excised for MS analysis. c , Top five candidate NPM1-interacting proteins identified by MS. d , IP and IB of the interaction between NPM1 and p65 in MNK3 cells. The samples were derived from the same experiment, and the blots were processed in parallel. e , Immunofluorescence staining of the subcellular location of p65 in ILC3s isolated from mice at day 5 of administration of 2.5% DSS or under the steady state (water). Scale bar = 5 μm. f , Unstimulated or stimulated MNK3 cells were subjected to cellular fractionation into cyto and nuc fractions followed by western blotting for p65. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and histone H3 were used as markers for cytosolic and nuclear proteins, respectively. g , RT–PCR analysis of mRNA abundance of p65 target genes in isolated colonic ILC3s from Npm1 + / + and Npm1 +/− mice at day 5 of administration of 2.5% DSS ( n  = 5 individual mice). h , KEGG pathway enrichment analysis of upregulated transcription factor-related pathways in ILC3s from patients with UC compared with ILC3s from healthy participants. i , Gene Ontology (GO) analysis of downregulated pathways in NPM1 low ILC3s compared with NPM1 high ILC3s from patients with UC, which was identified using a median expression cutoff for NPM1 in ILC3 of patients with UC. j – m , Cell mito stress test was performed with stimulated MNK3 cell line with or without p65 knockdown. Representative OCR ( j ), basal OCR ( k ), ATP production ( l ) and maximal respiration ( m ) are shown ( n  = 3 biological samples). n , Expression of Il22 in unstimulated and stimulated MNK3 cell line (shNC and sh p65 ; n  = 3 biological samples). Data in g and k – n are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t test (* P  < 0.05, ** P  < 0.01, *** P  < 0.001 and **** P  < 0.0001). IB, immunoblot; cyto, cytosol; nuc, nuclear.

To investigate whether NPM1 functions as a transcriptional cofactor that binds to p65 and influences the transcription of p65 target genes in activated ILC3s, we thus monitored the expression of four p65-regulated genes ( Cxcl2 , Ccl4 , Xiap and Cflar ) and found that they were dramatically induced in the colitis condition compared with the steady-state condition. However, only the expression of Cxcl2 and Ccl4 was markedly decreased in Npm1 +/− ILC3s compared to WT ILC3s from mice with DSS-induced colitis (Fig. 5g ). These transcriptional results indicated that the NF-κB pathway in ILC3s was activated by DSS-induced colitis and that NPM1 contributes to the regulation of a subset of NF-κB target genes. We also observed that the NF-κB signaling pathway in ILC3s from patients with UC was significantly upregulated compared to HCs in GSE182270 (Fig. 5h ). Additionally, several NF-κB-related signaling pathways were also enriched when comparing NPM1 high ILC3s and NPM1 low ILC3s from patients with UC (Fig. 5i ). In vitro tests of MNK3 cells with p65 knockdown exhibited a decrease in OCR, especially basal OCR (Fig. 5j–m ). More notably, the knockdown of p65 resulted in the downregulation of Il22 expression after stimulation (Fig. 5n ). Hence, our findings showed that p65 signaling was critical for the activation of ILC3. Meanwhile, NPM1 bound to p65 and participated in downstream transcriptional regulation in ILC3s in colitis.

To investigate a transcriptional regulatory role for NPM1 in mitochondrial OXPHOS and biogenesis, we examined the expression of the following three mitochondrial transcription factors in ILC3s: Tfam , mitochondrial transcription factor B1 ( Tfb1m ) and mitochondrial transcription factor B2 ( Tfb2m ). These transcription factors participate in mtDNA transcription and are stimulated by PGC1ɑ 45 . The expression levels of the three mitochondrial transcription factors in ILC3s showed no differences between Npm1 + / + and Npm1 +/− mice in steady state (Fig. 6a–c ). However, under pathological conditions, a remarkable decrease in Tfam expression was only observed in ILC3s, not macrophages, T cells and epithelial cells, of Npm1 +/− mice when compared to Npm1 + / + mice (Fig. 6a–c ), suggesting that NPM1 has an indispensable role in upregulation of Tfam in ILC3s upon DSS treatment. However, Tfb1m and Tfb2m were significantly increased in macrophages, T cells and epithelial cells, but not in ILC3s after DSS treatment (Fig. 6a–c ). These data suggested that mitochondrial activation in ILC3s is primarily dependent on TFAM rather than on TFB1M or TFB2M. Overexpression of Tfam in MNK3 markedly enhanced the expression of mtDNAs, including mt-Nd1, mt-Nd2, mt-Nd3 , mt-Nd4 and mt-Atp6 (Extended Data Fig. 8j ). Knockdown of Tfam in MNK3 notably impaired its mitochondrial function and attenuated the production of IL-22 (Fig. 6d–h ). Accordingly, NPM1 is crucial for the heightened demand of TFAM to subsequently increase mitochondrial function in ILC3s, not other cell types, during DSS-induced colitis.

a – c , RT–PCR analysis of mRNA expression of Tfam ( a ), Tfb1m ( b ) and Tfb2m ( c ) in isolated ILC3s, macrophages, T cells and epithelial cells from Npm1 + / + and Npm1 +/− mice exposed to 2.5% DSS or water (steady state; n  = 3 individual mice). d – g , Cell mito stress test was performed with stimulated MNK3 cell line with or without Tfam knockdown. Representative OCR ( d ), basal OCR ( e ), ATP production ( f ) and maximal respiration ( g ) are shown ( n  = 4 biological samples). h , Expression of Il22 in unstimulated and stimulated MNK3 cell line (shNC and sh Tfam ; n  = 3 biological samples). i , Logo plot of the consensus binding motif of the transcription factor p65. j , The positions and sequences of the four predicted binding sites of p65 in the TFAM promoter. k , Diagram of the pGL3- TFAM promoter luciferase reporter plasmids. l , TFAM reporter activity measured in HEK293T cells ( n  = 4 biological samples). m , ChIP–qPCR assays of the binding efficiency of p65 to the Tfam promoter in MNK3 cells with or without stimulation by IL-23 and IL-1β. IgG served as the negative control ( n  = 3 biological samples). n , o , Analysis of the effect of Tfam overexpression ( Tfam OE) on IL-22 production in MNK3 cells by flow cytometry ( n ) and ELISA ( o ), ( n  = 3 biological samples). p , Model depicting transcription activity change of Tfam in ILC3 cells with or without NPM1. Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-way ANOVA ( a – c , l , m ) and two-tailed unpaired Student’s t test ( e – h , o ; * P  < 0.05, ** P  < 0.01, *** P  < 0.001 and **** P  < 0.0001). TSS, transcription start site; shNC, nontargeted short hairpin RNA; sh Npm1 , short hairpin RNA targeting Npm1 .

To determine if NPM1 and p65 regulate TFAM expression, we examined whether they directly bind to the TFAM promoter and affect its transcription. We identified four putative binding sites for p65 in the TFAM promoter and constructed luciferase reporter plasmids (Fig. 6i–k ). Using luciferase reporter assays in HEK293T cells, we found that p65 significantly enhanced TFAM promoter-dependent reporter expression in plasmids with either third or fourth binding sites (Fig. 6l ). However, knockdown of NPM1 inhibited promoter activity (Fig. 6l ), indicating that NPM1 contributes to TFAM transcription.

To validate a direct interaction between p65 and the Tfam promoter in ILC3s, we performed chromatin immunoprecipitation (ChIP) using MNK3 cells and a p65 antibody and tested for the presence of Tfam promoter sequences. To determine the effect of inflammatory signals on the interaction, we evaluated MNK3 cells with and without stimulation by IL-1β and IL-23. We found that p65 is bound to the Tfam promoter in ILC3s under both conditions, with stimulation enhancing this interaction (Fig. 6m ). To confirm the importance of Tfam in ILC3 activation, we overexpressed Tfam in MNK3 cells and found that expression of Tfam mostly restored secretion of IL-22 in MNK3 cells in which Npm1 was knocked down (Fig. 6n,o ). Collectively, our findings indicated that NPM1 acts as a partner of p65 to promote Tfam transcription, thereby supporting ILC3 mitochondrial function and activation (Fig. 6p ).

Npm1 overexpression ( Npm1 OE) protects against DSS-induced colitis

Our subsequent investigation aimed to explain why NPM1 is downregulated in UC and whether overexpression of NPM1 could ameliorate colitis. GATA binding protein 3 (GATA3), interferon regulatory factor 1 (IRF1) and signal transducer and activator of transcription 3 (STAT3), which are predicted transcriptional factors associated with NPM1, demonstrated reduced expression in ILC3s of patients with UC and enteritic mice in comparison to the control groups (Extended Data Fig. 8k,l ). These findings may provide insights into the mechanisms underlying the downregulation of NPM1 in IBD. To confirm the protective function of NPM1 in colitis, we generated Npm1 UTR −/− mice that have a genetic knockout of the 3′-UTR region of Npm1 and overexpress Npm1 (Supplementary Fig. 1g–j ). Compared to control (Ctrl) mice, Npm1 UTR −/− mice had less severe DSS-induced colitis, as evidenced by the increased recovery of body weight, decreased DAI, increased colon length and reduced inflammation (Fig. 7a–e ). Although overexpression of Npm1 did not enhance the frequency of colonic ILC3s, a higher proportion were producing IL-22, indicating that Npm1 OE enhanced the defense function of ILC3s against colitis (Fig. 7f–j ). Expression of various mtDNA was upregulated in Npm1 UTR −/− ILC3 compared to control group (Extended Data Fig. 8m ). Overexpression of Npm1 in MNK3 cells also increased IL-22 secretion (Fig. 7k ). Eventually, we crossed Npm1 UTR −/− with Npm1 +/− mice and generated Npm1 UTR +/− Npm1 +/− mice. As expected, heterozygous overexpression of Npm1 prevented the exacerbated DSS-induced colitis caused by the Npm1 haploinsufficiency (Fig. 7l–n ). ILC3s isolated from Npm1 UTR +/− Npm1 +/− mice also showed increased IL-22 secretion compared with ILC3s from Npm1 +/− mice (Fig. 7o ). Taken together, these results demonstrated that Npm1 OE has a protective function against colitis.

a – e , Npm1 UTR −/− and littermate control mice were fed with 2.5% DSS for 7 days and allowed to recover for 3 days. Body weight ( a ), DAI ( b ), representative colon images ( c ), colon length ( d ) and colon histopathology ( e ) are shown ( n  = 4 individual mice). Scale bars = 500 μm (left) and 100 μm (right). f , g , LPLs were isolated from Npm1 UTR −/− and control mice on day 5 of administration of 2.5% DSS ( n  = 5 individual mice). The proportion of ILC3s (live CD45 + Lin − RORγt + cells) within the CD45 + population ( f ) and of IL-22-producing ILC3s (live CD45 + Lin − RORγt + IL-22 + cells) with the ILC3 population ( g ) was determined by flow cytometry. h , i , Number of ILC3s ( h ) and IL-22 + ILC3s ( i ) in LPLs of Npm1 UTR −/− and control mice after DSS administration are depicted ( n  = 5 individual mice). j , k , IL-22 production by isolated colonic ILC3s from Npm1 UTR −/− and control mice on day 5 of administration of 2.5% DSS ( j ) and MNK3 cells in response to Npm1 OE ( k ) was determined by ELISA ( n  = 3 biological samples). l – n , Control, Npm1 UTR +/− , Npm1 +/− and Npm1 UTR +/− Npm1 +/− mice were administered 2.5% DSS for 7 days and allowed to recover for 3 days. Representative images of the mouse colons ( l ), colon length ( m ) and colon histopathology ( n ) are shown ( n  = 5 individual mice). Scale bars = 500 μm (top) and 100 μm (bottom). o , IL-22 production by isolated colonic ILC3s from the indicated groups of mice was determined by ELISA ( n  = 3 individual mice). Data in d , f – k , m and o are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t test (* P  < 0.05, ** P  < 0.01 and *** P  < 0.001).

In this study, we demonstrated that NPM1, a protein that is abundant in colonic ILC3s, is critical for the activation of IL-22 production in response to colitis. We found that NPM1 binds to p65 and regulates transcription of the mitochondrial transcription factor TFAM , thereby having a role in maintaining mitochondrial biogenesis and OXPHOS. Our findings revealed the protective role of NPM1 in gut homeostasis and suggested that a deficiency in the activity of NPM1 is a key factor linking IBD and MDS/AML.

The NF-κB family of transcription factors has a crucial role in responding to various stimuli by regulating the expression of genes involved in diverse biological processes such as inflammation, metabolism, cancer and development 44 . Here we identified p65 as the top interacting protein with NPM1 in ILC3s in DSS-induced colitis and observed the subcellular translocation of p65 and colocalization with NPM1 in the nucleus of ILC3s to activate downstream gene transcription after inflammatory stimulation. A previous study revealed that NPM1 interacts with the N-terminal DNA-binding domain of p65 and enhances binding to target gene promoters 44 . We also found that TFAM is a potential target of p65 in ILC3s and that p65 regulates TFAM transcription in a manner enhanced by NPM1. TFAM is a mitochondrial transcription factor that controls mtDNA replication and transcription 46 . Tfam −/− mice are embryonically lethal, and tissue-specific deficiency of Tfam leads to severe OXPHOS defects, which is the main cause of human mitochondrial diseases 47 . Furthermore, Tfam ∆ILC3 mice exhibit a substantial reduction of ILC3s by 6 weeks of age 48 . Here we demonstrated that TFAM is highly expressed in ILC3s and acts as a key downstream effector of NPM1 in DSS-induced colitis. Moreover, mitochondrial activation in ILC3s is primarily dependent on TFAM, rather than on TFB1M or TFB2M. In contrast to ILC3s, macrophages, T cells and epithelial cells are primarily depend on TFB1M and/or TFB2M in DSS-treated mice. This indicates the indispensable role of NPM1 for ILC3s, not T cells, macrophages or epithelial cells. By maintaining mtDNA replication, mitochondrial number and OXPHOS levels in activated ILC3s, NPM1-stimulated TFAM expression supports the cells’ high energy requirements. Although there is no direct evidence that TFAM regulates IL-22, the lack of mitochondrial-derived energy by insufficient TFAM could limit the activation of ILC3s and effector cytokine secretion in colitis. Meanwhile, although it cannot be ruled out that NPM1 affects ILC3 mitochondrial function and cell activation through interactions with other molecules that participate in mitochondria functions, such as NPM1’s known partners c-Myc 49 , 50 , SP1 (refs. 51 , 52 ), p53 (refs. 53 , 54 ) and IRF1 (refs. 55 , 56 ), the effects of p65 knockdown and Tfam knockdown on ILC3 function in MNK3 cells are similar to those of knocking down Npm1 . Therefore, it is believed that the p65-TFAM axis is an important effector for NPM1 to increase the function and metabolism of ILC3.

Mitochondria have an important role in the activation of immune cells. Activation of ETC in mitochondria is essential for T cell activation, expansion and cytokine production 57 . The proliferation and cytokine secretion of ILC3s depend on glycolysis and also mitochondrial ROS following in vitro activation by IL-1β and IL-23 or in vivo during bacterial infection 39 . By analyzing mitochondria in primary ILC3s from Npm1 +/− mice under DSS administration, we observed diminished mitochondrial numbers, consistent with impaired biogenesis, and reduced OXPHOS. To confirm that the deficiency in IL-22 secretion was due to mitochondrial dysfunction, bezafibrate was used to activate mitochondrial function and successfully rescued the production of IL-22 in Npm1 -deficient ILC3s. Overexpression of Tfam also increased IL-22 production, providing additional support for mitochondrial biogenesis and OXPHOS were crucial for ILC3 activation.

MDS is a hematopoietic disorder involving clonal abnormalities of cells caused by mutations in oncogenes and tumor suppressor genes, as well as chromosomal abnormalities 9 . The subsequent alterations in the function and properties of BM-derived immune cells can lead to the development of immune-mediated disorders including IBD. Genetic mutations provide insight into the relationship between MDS and IBD. For example, mutations in PTPN11 , a driver gene in MDS/AML, result in exacerbation of intestinal inflammation by disrupting BM-derived macrophage responsiveness to IL-10 (ref. 58 ). NPM1 acts as a top driver mutation in high-risk MDS and AML 6 , 17 . In our study, we demonstrated that NPM1 functions in BM-derived ILC3s to control the gut microenvironment, particularly through a protective IL-22-related immune response. Our data provide insight with potential relevance for the diagnosis and treatment of patients with concurrent IBD and MDS.

In summary, our study highlights the role of NPM1 in maintaining mitochondrial function and IL-22 production in ILC3s in the progression of colitis. Our findings suggest that NPM1 might be a therapeutic target for IBD and provide insights into a connection between MDS/AML and IBD.

Generation of Npm1 +/− , Npm1 UTR +/− and Npm1 flox / + mice

Npm1 +/− and Npm1 UTR +/− mice were generated by knocking out the DNA-binding domain (including partial exon 8, exon 9, 10 and partial exon 11) and 3′-UTR domain with the binding sites of microRNAs using CRISPR–Cas9 technology from the CRO company Shanghai Model Organisms Center. In brief, Cas9 mRNA and gRNA were synthesized in vitro and then injected into fertilized eggs of C57BL/6J mice. The resulting F0 mice were screened for Npm1 +/− genotype using specific PCR primers (PI, 5′-GAAAAGGTCCCAGTGAAGAAAGTGA-3′; PII, 5′-TGGCAAGTGAACCTGGACAACAT-3′; PIII, 5′-GGCTGACCCACAGGCTGAGGAG-3′ and PIV, 5′-CCAACAGATTGGCTATCAATAGAGGA-3′) or Npm1 UTR +/− (PI, 5′-CCACAGGCTGAGGAGGCAACAC-3′; PII, 5′-AAAAGGTTCAGGCACGAAGCAG-3′; PIII, 5′-GTCAGATGTGGAAATGGTAGGGAGA-3′ and PIV, 5′-AAAAGGTTCAGGCACGAAGCAG-3′) and crossed with WT C57BL/6J mice to get F1 heterozygous mice which were identified by genotyping PCR. F1 heterozygous mice were crossed with WT C57BL/6J mice to get F2 heterozygous mice. The third and further generations of Npm1 +/− and Npm1 UTR +/− mice were used in the experiments.

Npm1 flox / + mice were generated by introducing two loxp sequences into Npm1 using CRISPR–Cas9 technology from the CRO company Cyagen Biosciences. Briefly, Cas9 mRNA and gRNA were synthesized in vitro with a homologous arms-encompassed targeting vector and injected into fertilized eggs of C57BL/6J mice. The resulting F0 mice were screened using specific PCR primers (FI, 5′-AACAGCTAGATGGGAAGTATGGA-3′; RI, 5′-AGTTCCCAAGTTTGCTTTGAACAG-3′ and FII, 5′-ACGTTGCAGATAGCTGTACTGATG-3′; RII, 5′-GCTAAAGCGAATCTTGTCTGTTCA-3′) and crossed with WT C57BL/6J mice to get F1 heterozygous mice, which were identified by genotyping PCR with primer pairs (F2 and R2). Positive F1 Npm1 flox / + mice were crossed with WT C57BL/6J mice to get F2 heterozygous mice. The third and further generations of Npm1 flox / + mice were used in the experiments.

All mice used in this study were bred in the animal facility of Suzhou Institute of Biomedical Engineering and Technology and Shandong University and were approved in accordance with the Institutional Animal Care and Use Committee guidelines at Suzhou Institute of Biomedical Engineering and Technology and Shandong University. Mice were housed in individually ventilated cages under a 12-h light/12-h dark cycle with normal food and water. All experiments were performed using C57BL/6J mice, which also served as controls for Npm1 +/− , Npm1 UTR −/− and Apc min /+ mice. Npm1 flox/flox mice served as controls for Rorc cre / + Npm1 flox / flox and Villin cre /+ Npm1 flox / flox mice. Male mice aged 6–8 weeks were used for the experiments. For the DSS model, drinking water containing 2.5% DSS was given to age-matched male mice for 7 days, followed by regular water for 3 days, with DSS water being replaced each day. For the rescue experiment, mice were treated with 10 mg kg −1 bezafibrate (i.g.) every other day. Throughout the experiment, body weight was monitored. To induce colon cancer model, WT and Npm1 +/− mice were injected intraperitoneally with AOM (10 mg kg −1 ). After 5 days, 2.5% DSS was added to the drinking water for seven consecutive days, followed by 14 days of regular water. This cycle was repeated three times. Mice were killed for analysis on day 65 of the experiments. In the TNBS model, mice were anesthetized and then treated with 2 mg of TNBS dissolved in 50% ethanol via rectal administration using a polyethylene catheter (2 mm in outer diameter). Following administration, the mice were maintained in an inverted position for a minimum of 1 min. Control mice were treated rectally with 50% ethanol alone. The progression of colitis was monitored daily, assessing parameters such as diarrhea, presence of blood in stools, body weight and survival rates. Note that littermate mice are generally genotyped at 3–4 weeks of age and then placed in separate cages when grouping, according to their genotype. The DAI is calculated by combining the following three parameters: the percentage weight loss of the mice, the consistency of stool and the presence of stool blood. The scoring for each parameter is as follows: (1) weight loss—0 points if weight remains stable, 1 point for a 1–5% weight loss, 2 points for a 5–10% weight loss, 3 points for a 10–15% weight loss and 4 points for a weight loss greater than 15%; (2) stool consistency—0 points for normal stool, 2 points for loose stool and 4 points for diarrhea and (3) stool blood—0 points for no blood, 2 points for occult blood positivity and 4 points for overt bleeding. The DAI is calculated as follows: DAI = (weight loss index + stool consistency + blood in stool)/3. Note that mice are generally genotyped and caged at 3–4 weeks of age and then placed in separate cages when grouping, according to their genotype.

Generation of BM chimera

The generation of BM chimeras was achieved by collecting BM cells from both WT and Npm1 +/− mice and subsequently flushing them with 1× PBS. The cell suspension, comprising 1 × 0 7 BM cells, was then intravenously injected into lethally irradiated recipient mice of both WT and Npm1 +/− genotypes, with a dose of 102.2 cGy min −1 for 9 min. Experiments were conducted 4 weeks following reconstitution.

In vivo T cell and myeloid cell blocking

To deplete T cells, anti-CD3ɛ (Bio X Cell, BE0001-1; clone 145-2C11) was administered intravenously daily (50 µg per mouse, from day −2 to day 6), and control mice were administered an equivalent amount of IgG (Bio X Cell, BE0091). To deplete myeloid cells, antimouse/antihuman CD11b (Bio X Cell, BE0007; clone M1/70) were administered intravenously every 2 days (100 µg per mouse, from day −2 to day 6), and control mice were administered an equivalent amount of IgG (Bio X Cell, BE0091; clone LTF-2). The DSS-induced colitis model was initiated on day 0.

We dissected the colons from the indicated mice, fixed them in 10% formalin and stained them with hematoxylin and eosin (H&E) using paraffin-embedded sections. We used the following scoring system to evaluate colon tissue histologically: 0 = no evidence of inflammation, 1 = low level of inflammation with scattered infiltrating mononuclear cells (1–2 foci), 2 = moderate inflammation with multiple foci, 3 = high level of inflammation with increased vascular density and marked wall thickening and 4 = maximal inflammation with transmural infiltration and loss of goblet cells.

Flow cytometry and isolation of lamina propria leukocytes

To isolate leukocytes from the lamina propria, we incubated intestinal segments of approximately 0.5 cm at 37 °C for 1.5 h in complete Roswell Park Memorial Institute (RPMI) medium (Suzhou Haixing Biosciences), supplemented with DNase I (150 µg ml −1 ; Sigma) and collagenase VIII (300 U ml −1 ; Sigma). The digested fragments were triturated and filtered through a 100 µm cell strainer. The cells were collected from the interface of the 80% and 40% Percoll gradients after centrifugation at 660 g for 15 min at room temperature. Before surface staining, Fc receptors were blocked using CD16/32 antibody (eBioscience; dilution 1:100). Leukocytes isolated from the intestinal lamina propria were then stained with antibodies against the following markers: CD45 eFlour 506 (dilution 1:100), RORγt PE (dilution 1:50), Ly-6G PE (dilution 1:100), CD127 Super Bright 645 (dilution 1:100), F4/80 FITC (dilution 1:100), CD3 Alexa-488 (dilution 1:100), CD34 FITC (dilution 1:100), CD117 APC (dilution 1:100), CD19 eFlour (450 dilution 1:100), IL-22 PE (dilution 1:50), CD4 APC (dilution 1:100), IL-17A BV421 (dilution 1:50), Lineage Percp-cy5.5 Cocktail (dilution 1:50), T-bet PE (dilution 1:100), IFNγ-APC (dilution 1:50), NKp46-PerCPcy5.5 (dilution 1:50), FOXP3-eFlour 450 (dilution 1:100), CCR6-BV421 (dilution 1:50), TCR γ/δ-APC (dilution 1:50) and CD127-FITC (dilution 1:100). For cytokine staining, cells were stimulated with phorbol 12-myristate 13-acetate (PMA) (50 ng ml −1 ) and ionomycin (500 ng ml −1 ) for 2 h, along with the addition of brefeldin A (2 µg ml −1 ). Live and dead cells were distinguished using the Live and Dead Violet Viability Kit (BioLegend).

Live + Lin − CD45 low CD90.2 high ILC3s were sorted from colon LPLs of the indicated mice. The SMARTer cDNA synthesis protocol was used to synthesize cDNA, which was then fragmented using dsDNA Fragmentase (New England Biolabs (NEB), M0348S) and incubated at 37 °C for 30 min. Library construction commenced with fragmented cDNA, where blunt-end DNA fragments were generated through a combination of fill-in reactions and exonuclease activity. Size selection was carried out using the provided sample purification beads. An A-base was added to the blunt ends of each strand, indexed Y adapters were ligated to the fragments and the ligated products were amplified using PCR. Subsequently, paired-end sequencing was conducted on NovaSeq 6000 (Illumina), following the protocol recommended by the vendor.

scRNA-seq data processing

scRNA-seq dataset ( GSE182270 ) was downloaded from the Gene Expression Omnibus (GEO) database and was performed on cells extracted from colonic biopsies of inflamed mucosa (patients with UC, n  = 5) and normal colonic mucosa (HCs, n  = 4). Count tables were analyzed using the Seurat 4.0 package following the standard workflow with default settings. The number of principal components (PCs) was determined based on Elbow plots, PCs = 13. Next, FindNeighbors and FindClusters functions were used for cell clustering, and the UMAP method was used for visualization. Cell-type-specific markers were found by the FindMarkers function; cell-type identities were manually annotated by matching cluster-specific upregulated marker genes with cell-type markers in the CellMarker 2.0 database. NPM1 low ILC3s and NPM1 high ILC3s were identified using a median expression cutoff for NPM1 in ILC3s. Note that the cell dropout of NPM1 was not included in the analysis. FindMarkers function was used to identify significantly regulated genes in NPM1 high ILC3. The ClusterProfiler package was applied for functional annotation.

Immunoprecipitation (IP) and western blot analysis

To perform IP, cells were lysed in an IP lysis buffer containing 20 mmol l −1 Tris (pH 7.5), 150 mmol l −1 NaCl and 1% Triton X-100, supplemented with a cocktail of protease and phosphatase inhibitors. Following lysis, the supernatants were collected after centrifugation and incubated overnight at 4 °C with constant rotation with the indicated antibodies. The antibody–antigen complexes were then precipitated using protein A/G magnetic beads (Millipore) and washed with PBS. For western blot analysis, cell lysates were prepared using radio immunoprecipitation assay (RIPA) lysis buffer (CoWin Biosciences) containing protease inhibitors and phosphatase inhibitors (CoWin Biosciences). Equal amounts of protein were loaded onto SDS–PAGE gels and transferred to nitrocellulose membranes. The membranes were blocked with 5% nonfat dried milk for 1 h at room temperature before being incubated with primary antibodies overnight at 4 °C, including SDHB (Proteintech; 1:2,000), NDUFB8 (Proteintech; 1:2,000), MT-ATP6 (Abclonal; 1:1,000), MT-CO1 (Abclonal; 1:1,000), UQCRC2 (Proteintech; 1:1,000), NPM1 (Abclonal; 1:1,000) and p65 (Cell Signaling Technology (CST); 1:1,000). After washing, the membranes were incubated with IRDye 800cw or 680cw conjugated secondary antibodies (LICORbio; 1:10,000) for 1 h. The membranes were then imaged using an Odyssey CLx Infrared Imaging System.

The ChIP assay was conducted using the ChIP-IT Kit (Beyotime). In brief, the cells were initially fixed with formaldehyde and subsequently lysed. To precipitate the DNA fragment, either 2 μg of anti-p65 or normal IgG were used. The DNA–protein complexes were then pulled down with magnetic beads and subjected to decross-linking. The extracted DNA samples were finally amplified using specific Tfam promoter primers for the sequences containing the binding site 5′-GGGAAAGGC-3′.

Luciferase reporter assay

HEK293T cells that overexpressed p65 were transfected with the specified pGL3-luciferase reporter plasmid that contained the TFAM promoter, along with the Renilla pRL-TK plasmid as the internal control. After incubation for 24–48 h, the cell lysates were subjected to luciferase activity analysis using the Dual-Luciferase Reporter Assay kit (Promega).

Immunofluorescence staining

Glass slides were pre-inserted into 12-well plates, and cells were seeded onto these plates. After 24 h, when the cells had reached 40–50% confluence, they were washed with PBS and subsequently fixed with 4% paraformaldehyde for 30 min. The cells were then permeabilized with a 0.5% Triton X-100 solution for an additional 20 min. A blocking buffer containing 5% bovine serum albumin was added next. Primary antibodies comprising anti-NPM1 (Proteintech; 1:200), anti-RORγt (Thermo Fisher Scientific; 1:200), anti-TOMM20 (Proteintech; 1:200) and anti-p65 (CST; 1:200) were used in this experiment. The corresponding secondary antibodies conjugated with Alexa Fluor 488, 555 and 647 were also used at a concentration of 1:2,000 (Invitrogen).

The Q Exactive Mass Spectrometer (Thermo Fisher Scientific) and Dionex Ultimate 3000 RSLCnano (Thermo Fisher Scientific) were used to analyze the affinity-purified samples according to the manufacturer’s instructions. The proteins were first reduced with 0.05 M Tris (2-carboxyethyl) phosphine (TCEP) and then alkylated with 55 mM methyl methanethiosulfonate (MMTS). The sample was then centrifuged and subjected to centrifugation steps before being digested with trypsin. After digestion, the resulting peptides were loaded onto a reversed-phase analytical column and underwent high-performance liquid chromatography (HPLC)–MS analysis. The peptide detection was conducted using an Orbitrap at a resolution of 70,000, with tandem mass spectrometry (MS/MS) using normalized collision energy (NCE) setting as 27, and MASCOT software was used to identify proteins. The peptide mass tolerance was 20 ppm, while the fragment mass tolerance was 0.6 Da, and the significance threshold was 0.05.

RNA extraction and quantitative real-time PCR analysis

The procedures were conducted as described previously 59 . Supplementary Table 2 lists all PCR primer sequences used for the detection of mt-Co1 , mt-Co2 , mt-Co3 , mt-Nd1 , mt-Nd2 , mt-Nd3 , mt-Nd4 , mt-Atp6 , Cxcl2 , Ccl4 , Xiap , cFlip , Tfam , Tfb1m , Tfb2m , Il22, Npm1, Gata3, IRF1, Stat3, Tjp1, Tjp2, Cldn2, Cldn3 and Gapdh . Moreover, Tfam ChIP–qPCR primers are also listed in Supplementary Table 2 .

The cells were centrifuged to precipitate, and the medium was removed. Then 500 μl of 2.5% glutaraldehyde (Ted Pella) was slowly added, avoiding suspending the precipitated cells, and left at room temperature for 1 h, followed by keeping at 4 °C for 3 h. The glutaraldehyde solution was replaced with PBS, and cells were left at 4 °C overnight. The cells were stained following the reported protocol 60 . The cells were then embedded in resin (Eponate 12 Kit, Ted Pella). Ultrathin sections of 50 nm thickness were cut (UC7, Leica) and collected on carbon-coated Kapton tapes. EM images were acquired with an SEM (GeminiSEM 300, Zeiss).

Mitochondrial membrane potential assay

Primary ILC3s were loaded with the JC-1 primer (Beyotime, C2006) and potentiometric dye TMRE (Beyotime, C2001S) at 37 °C for 20 min and washed with buffer or cell medium three times. Δ ψm was measured using a microplate reader. When detecting JC-1 monomers, the excitation light can be set to 490 nm and the emission light can be set to 530 nm. When detecting JC-1 polymer, the excitation light can be set to 525 nm and the emission light can be set to 590 nm. The maximum excitation wavelength of TMRE is 550 nm, and the maximum emission wavelength is 575 nm.

Immunohistochemical staining

Immunohistochemical staining was performed as follows: after deparaffinization and hydration, paraffin slides were repaired by boiling in Tris–EDTA buffer (pH 8.0) for 10 min. Next, sections were treated with 3% H 2 O 2 for 20 min to bleach endogenous peroxidase. After blocking with donkey serum, primary antibodies against NPM1 were diluted 1:100 and then incubated at 4 °C overnight. After three washes with PBS, the tissue slides were treated with horseradish peroxidase (HRP), conjugated donkey anti-rabbit or mouse secondary antibody (Dako) for 45 min and then stained by 3,3’-diaminobenzidine (DAB). Semi-quantitative immunohistochemistry is generally divided into the following three levels: low (+), medium (++) and high (+++). These levels are scored as follows: low (+) = 1, medium (++) = 2 and high (+++) = 3. Then calculate the value based on (+)% × 1 + (++)% × 2 + (+++)% × 3. The final score is (+) for a value less than 1.0, (++) for a value between 1.0 and 1.5 and (+++) for a value greater than 1.5.

Seahorse metabolic analysis

Cellular OCR was quantified using the Agilent Seahorse XFe24, following the manufacturer’s protocol. Primary ILC3s, macrophages, T cells, epithelial cells and MNK3 cells were plated on 24-well plates precoated with poly- d -lysine and incubated with the complete RPMI medium over night. Following the incubation period, the cells were washed and transferred into seahorse assay medium supplemented with 1 mM pyruvate, 2 mM glutamine and 10 mM glucose and cultured for an additional hour at 37 °C in a CO 2 -free environment. To measure OCR, indicated inhibitors such as oligomycin (1.5 μM), carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), 1 μM), rotenone (0.5 μM) and antimycin A (0.5 μM) were introduced where specified, and the rates of OCR (pmol O 2 per min) were monitored in real-time.

Pathology sections were obtained from patients with UC, patients with CD and healthy individuals after approval was obtained from the Ethics Committee of Shandong University School of Basic Medicine (ECSBMSSDU2020-1-035). All animal experiments were approved and are in accordance with the Institutional Animal Care and Use Committee guidelines at Suzhou Institute of Biomedical Engineering and Technology (2021-C058) and Shandong University (ECSBMSSDU2020-2-057).

Randomization and blinding

For DSS/TNBS/AOM-DSS animal studies to assess the changes in Npm1 deficiency, no method of randomization was used. Mice were grouped according to genotype, and all experiments were performed with sex-matched littermates. For the bezafibrate experiment, mice were first grouped by genotype and then randomly assigned to two groups (bezafibrate-treated group and control group). For CD11b/CD3 antibody treatment experiments, mice were first grouped by genotype and then randomly assigned to two groups (CD11b/CD3 antibody-treated group and IgG antibody-treated group). Animal studies were not blinded (mice were named with mouse ID and genotyped within 6 weeks of birth). Group allocation was not applicable because mice were grouped based on and compared across different genotypes. Histological analyses were conducted by two independent investigators, who had limited knowledge of the group of mice and patients. DAI was analyzed by two independent investigators, who had limited knowledge of the group of mice and patients. Data from FACS, qPCR and enzyme-linked immunosorbent assay (ELISA) were collected by an investigator with only the knowledge of mouse ID (without grouping information).

Statistical analysis

Flow cytometry data were analyzed by FlowJo (v10), and immunofluorescence images were analyzed by Image J 64-bit Java 8. Statistical analyses were performed using GraphPad Prism (v8). Data were presented as mean ± s.e.m. Statistical significance was assessed by Student’s t test (unpaired) or two-way analysis of variance (ANOVA) analyses. All statistical tests were two-tailed, and a P value <0.05 was considered statistically significant. Data distribution was assumed to be normal, but this was not formally tested. The number of patients, mice and biological repeats are indicated in the figure legends, as well as the number of independent experiments. No animal or data point was excluded from the analyses. In all studies using at least three to five animals per group, all experiments were performed at least twice to ensure reproducibility.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Smart-seq analysis of primary colonic ILC3 in Npm1 + / + and Npm1 +/− mice have been deposited in the GEO under the accession code GSE271455 . scRNA-seq data from patients with UC and healthy individuals was downloaded from GEO under the accession code GSE182270 . Source data are provided with this paper.

Neurath, M. Current and emerging therapeutic targets for IBD. Nat. Rev. Gastroenterol. Hepatol. 14 , 688 (2017).

Article   PubMed   Google Scholar  

Kobayashi, T. et al. Ulcerative colitis. Nat. Rev. Dis. Primers 6 , 74 (2020).

Baumgart, D. C. & Sandborn, W. J. Crohn’s disease. Lancet 380 , 1590–1605 (2012).

Danese, S. & Fiocchi, C. Ulcerative colitis. N. Engl. J. Med. 365 , 1713–1725 (2011).

Article   CAS   PubMed   Google Scholar  

Hibi, T. & Ogata, H. Novel pathophysiological concepts of inflammatory bowel disease. J. Gastroenterol. 41 , 10–16 (2006).

Ogawa, S. Genetics of MDS. Blood 133 , 1049–1059 (2019).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bejar, R., Levine, R. & Ebert, B. L. Unraveling the molecular pathophysiology of myelodysplastic syndromes. J. Clin. Oncol. 29 , 504–515 (2011).

Tefferi, A. & Vardiman, J. W. Myelodysplastic syndromes. N. Engl. J. Med. 361 , 1872–1885 (2009).

Kono, M. et al. Induction of complete remission by azacitidine in a patient with myelodysplastic syndrome-associated inflammatory bowel disease. J. Crohns Colitis 12 , 499–502 (2018).

Hebbar, M. et al. Association between myelodysplastic syndromes and inflammatory bowel diseases. Report of seven new cases and review of the literature. Leukemia 11 , 2188–2191 (1997).

Harewood, G. C., Loftus, E. V. Jr., Tefferi, A., Tremaine, W. J. & Sandborn, W. J. Concurrent inflammatory bowel disease and myelodysplastic syndromes. Inflamm. Bowel Dis. 5 , 98–103 (1999).

Wang, Z., Zhou, Y. & Liu, Y. Concurrent inflammatory bowel disease and myelodysplastic syndrome: report of nine new cases and a review of the literature. Dig. Dis. Sci. 53 , 1929–1932 (2008).

Nakamura, F. et al. Simultaneous occurrence of inflammatory bowel disease and myelodysplastic syndrome due to chromosomal abnormalities in bone marrow cells. Digestion 79 , 215–219 (2009).

Kristinsson, S. Y. et al. Chronic immune stimulation might act as a trigger for the development of acute myeloid leukemia or myelodysplastic syndromes. J. Clin. Oncol. 29 , 2897 (2011).

Article   PubMed   PubMed Central   Google Scholar  

Komrokji, R. S. et al. Autoimmune diseases and myelodysplastic syndromes. Am. J. Hematol. 91 , E280–E283 (2016).

Khan, N. et al. Incidence of acute myeloid leukemia and myelodysplastic syndrome in patients with inflammatory bowel disease and the impact of thiopurines on their risk. Am. J. Gastroenterol. 116 , 741–747 (2021).

Makishima, H. et al. Dynamics of clonal evolution in myelodysplastic syndromes. Nat. Genet. 49 , 204–212 (2017).

Hingorani, K., Szebeni, A. & Olson, M. O. Mapping the functional domains of nucleolar protein B23. J. Biol. Chem. 275 , 24451–24457 (2000).

Uckelmann, H. J. et al. Mutant NPM1 directly regulates oncogenic transcription in acute myeloid leukemia. Cancer Discov. 13 , 746–765 (2023).

Wang, X. Q. D. et al. Mutant NPM1 hijacks transcriptional hub to maintain pathogenic gene programs in acute myeloid leukemia. Cancer Discov. 13 , 724–745 (2023).

Morganti, C. et al. NPM1 ablation induces HSC aging and inflammation to develop myelodysplastic syndrome exacerbated by p53 loss. EMBO Rep. 23 , e54262 (2022).

Raval, A. et al. Effect of nucleophosmin 1 haploinsufficiency on hematopoietic stem cells. Leukemia 26 , 853–855 (2012).

Artis, D. & Spits, H. The biology of innate lymphoid cells. Nature 517 , 293–301 (2015).

Eberl, G. et al. An essential function for the nuclear receptor RORγt in the generation of fetal lymphoid tissue inducer cells. Nat. Immunol. 5 , 64–73 (2004).

Kiss, E. A. et al. Natural aryl hydrocarbon receptor ligands control organogenesis of intestinal lymphoid follicles. Science 334 , 1561–1565 (2011).

Qiu, J. et al. The aryl hydrocarbon receptor regulates gut immunity through modulation of innate lymphoid cells. Immunity 36 , 92–104 (2012).

Guo, X. et al. Induction of innate lymphoid cell-derived interleukin-22 by the transcription factor STAT3 mediates protection against intestinal infection. Immunity 40 , 25–39 (2014).

Sonnenberg, G. F. et al. Innate lymphoid cells promote anatomical containment of lymphoid-resident commensal bacteria. Science 336 , 1321–1325 (2012).

Huber, S. et al. IL-22BP is regulated by the inflammasome and modulates tumorigenesis in the intestine. Nature 491 , 259–263 (2012).

Uzzan, M. et al. Ulcerative colitis is characterized by a plasmablast-skewed humoral response associated with disease activity. Nat. Med. 28 , 766–779 (2022).

Schinke, C. et al. IL8-CXCR2 pathway inhibition as a therapeutic strategy against MDS and AML stem cells. Blood 125 , 3144–3152 (2015).

Muto, T. et al. Concurrent loss of Ezh2 and Tet2 cooperates in the pathogenesis of myelodysplastic disorders. J. Exp. Med. 210 , 2627–2639 (2013).

Gobert, A. P. et al. Protective role of spermidine in colitis and colon carcinogenesis. Gastroenterology 162 , 813–827 (2022).

Shah, S. C. & Itzkowitz, S. H. Colorectal cancer in inflammatory bowel disease: mechanisms and management. Gastroenterology 162 , 715–730 (2022).

Vu, T. et al. Serine-threonine kinase receptor-associated protein is a critical mediator of APC mutation-induced intestinal tumorigenesis through a feed-forward mechanism. Gastroenterology 162 , 193–208 (2022).

Kuo, W. T. et al. The tight junction protein ZO-1 is dispensable for barrier function but critical for effective mucosal repair. Gastroenterology 161 , 1924–1939 (2021).

Luoma, A. M. et al. Molecular pathways of colon inflammation induced by cancer immunotherapy. Cell 182 , 655–671 (2020).

Allan, D. S. et al. An in vitro model of innate lymphoid cell function and differentiation. Mucosal Immunol. 8 , 340–351 (2015).

Di Luccia, B., Gilfillan, S., Cella, M., Colonna, M. & Huang, S. C.-C. ILC3s integrate glycolysis and mitochondrial production of reactive oxygen species to fulfill activation demands. J. Exp. Med. 216 , 2231–2241 (2019).

Xiong, L. et al. Nutrition impact on ILC3 maintenance and function centers on a cell-intrinsic CD71–iron axis. Nat. Immunol. 24 , 1671–1684 (2023).

Chamoto, K. et al. Mitochondrial activation chemicals synergize with surface receptor PD-1 blockade for T cell-dependent antitumor activity. Proc. Natl Acad. Sci. USA 114 , E761–E770 (2017).

Inak, G. et al. Defective metabolic programming impairs early neuronal morphogenesis in neural cultures and an organoid model of Leigh syndrome. Nat. Commun. 12 , 1929 (2021).

Weinberg, S. E., Sena, L. A. & Chandel, N. S. Mitochondria in the regulation of innate and adaptive immunity. Immunity 42 , 406–417 (2015).

Lin, J., Kato, M., Nagata, K. & Okuwaki, M. Efficient DNA binding of NF-κB requires the chaperone-like function of NPM1. Nucleic Acids Res. 45 , 3707–3723 (2017).

PubMed   Google Scholar  

Barshad, G., Marom, S., Cohen, T. & Mishmar, D. Mitochondrial DNA transcription and its regulation: an evolutionary perspective. Trends Genet. 34 , 682–692 (2018).

Wculek, S. K. et al. Oxidative phosphorylation selectively orchestrates tissue macrophage homeostasis. Immunity 56 , 516-+ (2023).

Ekstrand, M. I. et al. Mitochondrial transcription factor A regulates mtDNA copy number in mammals. Hum. Mol. Genet. 13 , 935–944 (2004).

Fu, Z. et al. Mitochondrial transcription factor A in RORγt + lymphocytes regulate small intestine homeostasis and metabolism. Nat. Commun. 12 , 4462 (2021).

Lee, K. M. et al. MYC and MCL1 cooperatively promote chemotherapy-resistant breast cancer stem cells via regulation of mitochondrial oxidative phosphorylation. Cell Metab. 26 , 633–647 (2017).

Li, Z. & Hann, S. R. Nucleophosmin is essential for c-Myc nucleolar localization and c-Myc-mediated rDNA transcription. Oncogene 32 , 1988–1994 (2013).

Scarpulla, R. C. Transcriptional paradigms in mammalian mitochondrial biogenesis and function. Physiol. Rev. 88 , 611–638 (2008).

Dhar, S. K., Xu, Y., Chen, Y. M. & Clair, D. K. S. Specificity protein 1-dependent p53-mediated suppression of human manganese superoxide dismutase gene expression. J. Biol. Chem. 281 , 21698–21709 (2006).

Wang, H. et al. p53 contributes to cardiovascular diseases via mitochondria dysfunction: a new paradigm. Free Radic. Biol. Med. 208 , 846–858 (2023).

Ånensen, N. et al. Correlation analysis of p53 protein isoforms with mutations and therapy response in acute myeloid leukemia. Oncogene 31 , 1533–1545 (2012).

Zhong, Z. Y. et al. New mitochondrial DNA synthesis enables NLRP3 inflammasome activation. Nature 560 , 198–203 (2018).

Kondo, T. et al. Identification and characterization of nucleophosmin/B23/numatrin which binds the anti-oncogenic transcription factor IRF-1 and manifests oncogenic activity. Oncogene 15 , 1275–1281 (1997).

Ron-Harel, N. et al. Mitochondrial biogenesis and proteome remodeling promote one-carbon metabolism for T cell activation. Cell Metab. 24 , 104–117 (2016).

Xiao, P. et al. Phosphatase Shp2 exacerbates intestinal inflammation by disrupting macrophage responsiveness to interleukin-10. J. Exp. Med. 216 , 337–349 (2019).

Zhou, Y. et al. USF1/CD90 signaling in maintaining glioblastoma stem cells and tumor-associated macrophages adhesion. Neuro Oncol. 24 , 1482–1493 (2022).

Hua, Y., Laserstein, P. & Helmstaedter, M. Large-volume en-bloc staining for electron microscopy-based connectomics. Nat. Commun. 6 , 7923 (2015).

Download references

Acknowledgements

We thank Z. Zhang’s laboratory, W. Feng and H. Zhu for their help and suggestions. Advanced Optical Microscopy Application Open Platform in Suzhou Institute of Biomedical Engineering and Technology (SIBET) is acknowledged for professional assistance in image acquisition. We thank N.R. Gough (BioSerendipity) for professional editing services. This work was funded by the Chinese Academy of Sciences (CAS) Project for Young Scientists in Basic Research (grant YSBR-067 to G.S.), the Basic Research Pilot Program of Suzhou (SJC2022007 to M.S.), National Natural Science Foundation of China (82273336 to M.S. and 82071854 to S.L), Pre-Research and Construction Project of Major Scientific Research Facilities in Jiangsu Province (BM2022010 to M.S.), National Key R&D Program of China (2020YFA0804400 to S.L.) and Innovative Research Group Project (82321002 to S.L.).

Author information

These authors contributed equally: Rongchuan Zhao, Jiao Yang.

Authors and Affiliations

Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Science, Suzhou, China

Rongchuan Zhao, Hong Zhang, Yuanshuai Zhou, Lei Hong, Jinlin Pan, Shaheryar Shafi, Guohua Shi, Ruobing Zhang, Dingsan Luo, Jinyun Yuan & Minxuan Sun

School of Biomedical Engineering (Suzhou), Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei, China

Rongchuan Zhao, Yuanshuai Zhou, Lei Hong, Jinlin Pan, Shaheryar Shafi, Guohua Shi, Ruobing Zhang & Minxuan Sun

Suzhou Hospital, Affiliated Hospital of Medical School, Nanjing University, Suzhou, China

Jiao Yang & Yanxiang Liu

Advanced Medical Research Institute, Shandong University, Jinan, China

Yunjiao Zhai & Shiyang Li

Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, China

Detian Yuan

The First Rehabilitation Hospital of Shanghai, Brain and Spinal Cord Innovation Research Center, School of Medicine, Advanced Institute of Translational Medicine, Tongji University, Shanghai, China

Ruilong Xia & Changgeng Peng

CAM-SU Genomic Resource Center, Soochow University, Suzhou, China

Shanghai Key Laboratory of Anesthesiology and Brain Functional Modulation, Translational Research Institute of Brain and Brain-Like Intelligence, Shanghai Fourth People’s Hospital, School of Medicine, Tongji University, Shanghai, China

Changgeng Peng

You can also search for this author in PubMed   Google Scholar

Contributions

R.C.Z., J.Y., S.Y.L. and M.X.S. conceptualized the project. R.C.Z. and J.Y. developed the methodology. R.C.Z., Y.J.Z., H.Z., J.L.P., L.H., D.S.L. and J.Y.Y. performed validation. Y.S.Z. and D.T.Y. carried out a formal analysis. R.C.Z., Y.J.Z., H.Z. and L.H. conducted the investigation. M.X.S., S.Y.L., Y.X.L., D.J.P., R.L.X. and C.G.P. arranged the resources. R.C.Z., Y.S.Z. and D.T.Y. curated the data. R.C.Z. wrote the original draft. J.Y., S.S., S.Y.L., D.T.Y., C.G.P. and M.X.S. did the writing, reviewing and editing of the manuscript. M.X.S., S.Y.L., C.G.P., R.B.Z. and G.H.S. provided supervision.

Corresponding authors

Correspondence to Changgeng Peng , Shiyang Li or Minxuan Sun .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Immunology thanks Matthew Hepworth and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Nick Bernard, in collaboration with the Nature Immunology team. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended data fig. 1 npm1 deficiency increases susceptibility to enteritis and colonic adenocarcinoma..

( a ) UMAP analysis of GSE182270 shows different clusters of cells in human colonic biopsies. ( b ) Expression of NPM1 in different clusters (y axis). Dot size represents the fraction of cells within the cluster that express each NPM1 . Colors indicate the z-scaled expression of genes in cells within each cluster. ( c ) Expression of NPM1 in different stages of 425 COAD patients in TCGA. (Stages I and II: n = 238 individual patients; stages III and IV: n = 187 individual patients). ( d ) Survival analysis of NPM1 in 270 COAD patients in TCGA, cutoff by the median expression of NPM1 in groups. ( e ) Survival analysis of NPM1 in 92 READ patients in TCGA, cutoff by the median expression of NPM1 in groups. ( f , g ) Protein expression analysis of NPM1 in the colon of Npm1 + / + and Npm1 +/− mice. Quantitative analysis of protein levels of NPM1, relative to tubulin ( e ) (n = 3 individual mice). ( h ) RT-PCR analysis of mRNA expression of Npm1 in whole colon of Npm1 + / + and Npm1 +/− mice (n = 7 individual mice). ( i ) Representative images of colons from Npm1 + / + and Npm1 +/− mice in steady state. ( j ) H&E staining of colon from Npm1 + / + and Npm1 +/− mice in steady state. Scale bars: 100 μm. ( k ) Comparison of mesenteric lymph nodes from Npm1 + / + and Npm1 +/− mice in steady state (n = 4). ( l , m ) Representative images of Peyer’s patches from Npm1 + / + and Npm1 +/− mice in steady state ( l ). Analysis of the number of Peyer’s patches ( m ) was performed (n = 5 individual mice). ( n ) H&E staining of solitary intestinal lymphoid tissue from Npm1 + / + and Npm1 +/− mice in steady state. Scale bars: 100 μm. ( o ) Ratio of LK, LSK, Lin − Sca1 low CD117 low cells ( Npm1 + / + : n = 7 individual mice; Npm1 +/− : n = 6 individual mice), CMP, GMP and MEP ( Npm1 + / + : n = 5 individual mice; Npm1 +/− : n = 6 individual mice) in bone marrow from Npm1 + / + and Npm1 +/− mice under steady-state. ( p ) Ratio of LK, LSK, Lin − Sca1 low CD117 low cells, CMP, GMP and MEP in bone marrow from Npm1 + / + and Npm1 +/− mice under steady-state (n = 6 individual mice). ( q – u ) Npm1 +/− and control Npm1 + / + mice were administered trinitrobenzene sulfonic acid (TNBS). Colon length ( q , r ), colon histopathology on day 7 ( s ), body weight ( t ) and DAI ( u ) were analyzed (n = 5 individual mice). Scale bars: 500 μm (up), 100 μm (down). ( v , w ) Npm1 +/− and control Npm1 + / + mice were treated with AOM-DSS for 65 days, and body weight ( v ) and DAI ( w ) were analyzed (n = 5 individual mice). ( x ) Representative images of colons from CRC mouse model. Data in c , g , h , m , o , p and r are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined two-tailed unpaired Student’s t -test (* p  < 0.05, ** p  < 0.01 and *** p  < 0.001).

Extended Data Fig. 2 NPM1 in hematopoietic system is essential to colon against colitis.

( a ) RT-PCR analysis of mRNA abundance of Npm1 in ILC3s, macrophages, T cells and epithelial cells from Npm1 + / + and Npm1 +/− mice exposed to 2.5% DSS or water (steady state) (n = 3 individual mice). ( b – f ) Bone marrow chimeric mice of indicated genotypes were treated with 2.5% DSS water for 7 days, and representative images of the mouse colons on day 10 of the DSS model ( b ), body weight ( c ), colon length ( d ) and histopathology ( e , f ) were analyzed (n = 3 individual mice). Scale bars represent 100 μm. ( g – j ) RT-PCR analysis of mRNA expression of the indicated genes in epithelial cells from Npm1 + / + and Npm1 +/− mice exposed to 2.5% DSS or water (steady state) (n = 3 individual mice). ( k – m ) Villin cre / + Npm1 +/− and control Npm1 flox / flox mice were treated with 2.5% DSS for 7 days, and representative images ( k ), colon length ( l ) and histopathology ( m ) were analyzed (n = 5 individual mice). Scale bars: 500 μm (left), 100 μm (right). Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-way ANOVA ( a ) and two-tailed unpaired Student’s t -test ( d , f – j , l ) (* p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001).

Extended Data Fig. 3 Regulation of colitis by NPM1 is independent of myeloid cells and T cells.

( a – h ) Proportion of eosinophil, macrophage, neutrophil and dendritic cells (DCs) in lamina propria lymphocytes (LPLs) of Npm1 + / + and Npm1 +/− mice under steady-state conditions ( a – d ) ( Npm1 + / + : n = 5 individual mice; Npm1 +/− : n = 6 individual mice) and during DSS-induced colitis ( e – h ) (n = 6 individual mice) state are shown. ( i – m ) Colitis in N pm1 + / + and Npm1 +/− mice was induced by DSS following administration with IgG or anti-CD11b blocking antibody. Deletion of CD11b+ cells by antibody ( i ). Representative images of colons ( j ), colon length ( k ) (n = 5 individual mice), colon histopathology on day 10 ( l ) and DAI ( m ) (n = 3 individual mice) are presented. Mice were injected with CD11b antibody (100 μg per mouse) every 2 days (from day −2 to day 6). Scale bars: 500 μm (up), 100 μm (down). ( n – s ) Proportion of T H 17, T reg and γδT in LPLs of Npm1 + / + and Npm1 +/− mice under steady-state conditions ( n – p ) ( Npm1 + / + : n = 5 individual mice; Npm1 +/− : n = 6 individual mice) and during DSS-induced colitis ( q – s ) (n = 5 individual mice) state are shown. ( t – x ) DSS-induced colitis in N pm1 + / + and Npm1 +/− mice was established following administration with IgG or anti-CD3 blocking antibody. Deletion of CD3 + cells by antibody ( t ). Representative images of colons ( u ), colon length ( v ) (n = 5 individual mice), colon histopathology ( w ) and DAI ( x ) (n = 3 individual mice) on day 10 are presented. Mice were injected with CD3 antibody (50 μg per mouse) once a day (from day −2 to day 6). Scale bars: 500 μm (up), 100 μm (down). Data in a – h , k , m , n – s , v and x are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined two-tailed unpaired Student’s t -test (** p  < 0.01 and **** p  < 0.0001).

Extended Data Fig. 4 Changes of ILC3 may contribute to dysbiosis of fecal microflora under pathological conditions.

( a – d ) LPLs were isolated from Npm1 + / + and Npm1 +/− mice on day 5 after administration of TNBS. ( a ) Proportion of ILC3s in Lin − cells. ( b ) Proportion of IL-22 + ILC3s in the total ILC3 population. Number of ILC3s ( c ) and IL-22 + ILC3s ( d ) are shown ( Npm1 + / + : n = 5 individual mice; Npm1 +/− : n = 4 individual mice). ( e – h ) ILC3s in LPLs of Npm1 + / + and Npm1 +/− mice under steady-state. ( e ) Proportion of ILC3s in Lin − cells. ( f ) Proportion of IL-22 + ILC3s in the total ILC3 population (n = 6 individual mice). Number of ILC3s ( g ) and IL-22 + ILC3s ( h ) are depicted ( Npm1 + / + : n = 6 individual mice; Npm1 +/− : n = 5 individual mice). ( i , j ) Proportion of NCR + ILC3s and CCR6 + ILC3s in total ILC3s from Npm1 + / + and Npm1 +/− mice under steady-state ( i ) (n = 5 individual mice) and during DSS-induced colitis ( j ) (n = 6 individual mice). ( k – q ) Feces from Npm1 + / + mice and Npm1 +/− mice under colitis were collected to analyze intestinal microbiota by 16S rRNA sequencing. ( k ) Observed operational taxonomic unit (OTU), ( l ) Chao1 index, ( m ) Shannon–Wiener diversity index (Shannon index) and ( n ) principal coordinates analysis (PCoA). ( o ) Venn diagram of two groups of fecal microbiota. ( p ) The relative abundance of microbiota at phylum level in the fecal samples. ( q ) Heatmap analysis of the relative abundance of microbiota at family level in the fecal samples. (n = 6 individual mice). ( r – u ) Co-housed Npm1 + / − and control Npm1 + / + mice were administered 2.5% DSS for 7 days, followed by 3 days of recovery (H 2 O). Body weight ( r ), DAI ( s ) and colon length on day 10 ( t , u ) were analyzed (n = 5 individual mice). Scale bars: 500 μm (left), 100 μm (right). Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t -test ( a – h , k – m , u ) and two-way ANOVA ( i , j ) (* p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001).

Extended Data Fig. 5 NPM1 is required for maintaining the frequency and function of colonic ILC3s.

( a ) Images of mesenteric lymph nodes from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice (n = 4 individual mice). ( b , c ) Representative images of Peyer’s patches from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice in steady state. The number of Peyer’s patches ( c ) was analyzed (n = 5 individual mice). ( d – g ) ILC3s in LPLs of Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice under steady-state. ( d ) Proportion of ILC3s in Lin − cells. ( e ) Proportion of IL-22 + ILC3s in the total ILC3 population. Number of ILC3s ( f ) and IL-22 + ILC3s ( g ) are depicted (n = 5 individual mice). ( h ) Images of mesenteric lymph nodes from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice (n = 3 individual mice). ( i , j ) Representative images of Peyer’s patches from Npm1 flox / + and Rorc cre / + Npm1 flox / + mice in steady state ( i ). The number of Peyer’s patches ( j ) was analyzed (n = 5 individual mice). ( k – m ) Npm1 flox / + and Rorc cre / + Npm1 flox / + mice were administered 2.5% DSS for 7 days followed by 3 days of recovery. Representative images of colons ( k ), colon length ( l ) and colon histopathology on day 10 ( m ) are presented. (n = 4 individual mice). Scale bars: 500 μm (up), 100 μm (down). ( n – q ) ILC3s in LPLs of Npm1 flox / + and Rorc cre / + Npm1 flox / + mice with colitis. ( n ) Proportion of ILC3s in Lin − cells. ( o ) Proportion of IL-22 + ILC3s in the total ILC3 population. Number of ILC3s ( p ) and IL-22 + ILC3s ( q ) are provided (n = 4 individual mice). ( r , s ) Apoptotic percentage of ILC3s in LPLs from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice was detected by Annexin V staining under steady-state ( r ) and during DSS-induced colitis ( s ) (n = 5 individual mice). ( t , u ) Proportion of NCR + ILC3s and CCR6 + ILC3s in total ILC3s from colon of Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice under steady-state ( t ) and during DSS-induced colitis ( u ) (n = 5 individual mice). ( v – y ) Proportion of T-bet + ILC3s in LPLs and IFNγ + cells in T-bet + ILC3s from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice under steady-state ( v , w ) (n = 4 individual mice) and during DSS-induced colitis ( x , y ) (n = 5 individual mice). Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t -test ( c – g , j , l , n – q and v – y ) and two-way ANOVA ( r – u ) (* p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001).

Extended Data Fig. 6 Exacerbate enteritis in Rorc cre / + Npm1 flox / flox mice under DSS is independent of T cells.

( a – n ) Proportions of eosinophil, dendritic cells (DCs), macrophages, neutrophils, T H 17, T reg and γδT in LPLs from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice under steady-state ( a – d , i – k ) and during DSS-induced colitis ( e – h , l – n ) (n = 5 individual mice). ( o – q ) Colitis in Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice was induced by DSS following administration with IgG or anti-CD3 blocking antibody. Representative images of colons ( o ), colon length ( p ) and colon histopathology ( q ) on day 10 are presented (n = 5 individual mice). Mice were injected with CD3 antibody (50 μg per mouse) once a day (from day −2 to day 6). Scale bars: 500 μm (up), 100 μm (down). ( r ) Representative images of colons with tumors from Npm1 flox / flox and Rorc cre / + Npm1 flox / flox on day 65 of the AOM/DSS CAC model. ( s , t ) Total number of tumors ( s ) and number of tumors larger than 2 mm ( t ) in Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice (n = 4 individual mice). Data in a – n , p , s and t are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined two-tailed unpaired Student’s t -test (* p  < 0.05, ** p  < 0.01 and *** p  < 0.001).

Extended Data Fig. 7 NPM1 is essential for mitochondrial function in ILC3s.

( a – c ) RT-PCR analysis of mRNA abundance of the indicated genes in macrophage ( a ), T cells ( b ) and epithelial ( c ), sorted from the LPLs of Npm1 + / + and Npm1 + / − mice in steady state or at day 5 of administration of 2.5% DSS (n = 3 individual mice). ( d – g ) Cell mito stress test was performed with sorted colonic ILC3s from Npm1 + / + and Npm1 + / − mice under steady-state. Representative oxygen consumption rate profile (OCR) ( d ), basal OCR ( e ), ATP production ( f ) and maximal respiration ( g ) are shown (n = 3 individual mice). ( h – k ) Cell mito stress test was conducted with sorted colonic ILC3s from wild-type (WT) mice on days 0, 5 and 10 of DSS-induced colitis. Representative OCR ( h ), basal OCR ( i ), ATP production ( j ) and maximal respiration ( k ) are presented (n = 3 individual mice). ( l – t ) Cell mito stress test was performed with sorted colonic macrophages ( l – n ), T cells ( o – q ), epithelial cells ( r – t ) from LPLs of Npm1 + / + and Npm1 + / − mice at day 5 of administration of 2.5% DSS (n = 4 individual mice). ( u , v ) Mitochondrial membrane potential was assessed with the indicator JC-1 ( u ) and TMRE ( v ) in isolated colonic ILC3s from Npm1 + / + and Npm1 + / − mice in steady state (n = 5 individual mice). ( w – y ) Npm1 flox / flox and Rorc cre / + Npm1 flox / flox mice were treated with bezafibrate (i.g., 10 mg/kg) and administered 2.5% DSS for 7 days followed by 3 days of recovery. Representative colon images ( w ), colon length ( x ) and colon histopathology ( y ) are shown (n = 5 individual mice). Scale bars: 500 μm (up), 100 μm (down). Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-way ANOVA ( a – c ) and two-tailed unpaired Student’s t -test ( e – g , i – v , x ) (* p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001).

Extended Data Fig. 8 Adequate mitochondrial function is critical to ILC3 activation.

( a – d ) Cell mito stress test was performed with unstimulated and stimulated MNK3 cell lines (shNC and sh Npm1 ). Representative oxygen consumption rate profile (OCR) ( a ), basal OCR ( b ), ATP production ( c ) and maximal respiration ( d ) are shown (n = 3 biological samples). ( e , f ) MNK3 cells with or without Npm1 -knockdown were treated with bezafibrate, and the MFI of IL-22 in MNK3 was analyzed by FC (n = 3 biological samples). ( g – i ) MNK3 cells with or without Npm1 -knockdown were treated with oligomycin ( g ), rotenone ( h ) and succinate ( i ), and the expression of Il22 was analyzed by RT-PCR (n = 5 biological samples). ( j ) RT-PCR analysis of mRNA abundance of the indicated genes in control and Tfam OE MNK3 cells (n = 4 biological samples). ( k ) Expression of indicated genes in ILC3s of UC and HC in GSE182270 . ( l ) RT-PCR analysis of mRNA abundance of Gata3 (n = 5 indicated mice), Irf1 (n = 3 indicated mice), Stat3 (n = 3 indicated mice) in ILC3s of mice in steady state or in DSS-induced colitis. ( m ) RT-PCR analysis of mRNA abundance of the indicated genes in wild-type and Npm1 UTR −/− ILC3s at day 5 of administration of 2.5% DSS (n = 4 indicated mice). Data are representative of two independent experiments, shown as the means ± s.e.m., and statistical significance was determined by two-tailed unpaired Student’s t -test ( b – d , f – i and l ) and two-way ANOVA ( j , m ) (* p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001).

Supplementary information

Supplementary information.

Supplementary Figs. 1 and 2 and Supplementary Tables 1 and 2.

Reporting Summary

Peer review file, source data fig. 1.

Statistical source data.

Source Data Fig. 2

Source data fig. 3.

Unprocessed western blots.

Source Data Fig. 4

Source data fig. 5, source data fig. 6, source data fig. 7, source data extended data fig. 1, source data extended data fig. 2, source data extended data fig. 3, source data extended data fig. 4, source data extended data fig. 5, source data extended data fig. 6, source data extended data fig. 7, source data extended data fig. 8, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Zhao, R., Yang, J., Zhai, Y. et al. Nucleophosmin 1 promotes mucosal immunity by supporting mitochondrial oxidative phosphorylation and ILC3 activity. Nat Immunol (2024). https://doi.org/10.1038/s41590-024-01921-x

Download citation

Received : 12 June 2023

Accepted : 10 July 2024

Published : 05 August 2024

DOI : https://doi.org/10.1038/s41590-024-01921-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Cancer newsletter — what matters in cancer research, free to your inbox weekly.

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Design and experiment of planting mechanism of automatic transplanter for densely planted vegetables.

1. Introduction

2. device structure and working principle, 2.1. structure and working principle of transplanter, 2.2. structure and working principle of planting mechanism, 3. design of key components, 3.1. establishment of kinematic model of planting mechanism, 3.2. design of power transmission mechanism, 3.2.1. design of motor and screw, 3.2.2. design of mechanical-pneumatic coupling duckbill opening and closing mechanism, 3.3. structural design of duckbill planting end effector, 3.3.1. strength analysis of linkage rod of planting mechanism, 3.3.2. coupling simulation test of duckbill planting end effector, 4. resistance verification test of duckbill planting end effector, 4.1. test design, 4.2. analysis of test results, 5. planting performance test, 5.1. test conditions, 5.2. test design and evaluation index, 5.3. test results and analysis, 6. conclusions, author contributions, institutional review board statement, data availability statement, conflicts of interest.

• Periasamy, V.; Gounder, D.V.M.; Ramasamy, K. Factors influencing the performance of mechanical end effector during automatic transplanting of tomato seedlings. J. Appl. Nat. Sci. 2022 , 14 , 227–231. [ Google Scholar ] [ CrossRef ]
• Wu, G.; Wang, S.; Zhang, A.; Xiao, Y.; Li, L.; Yin, Y.; Li, H.; Wen, C.; Yan, B. Optimized design and experiment of a self-covering furrow opener for an automatic sweet potato seedling transplanting machine. Sustainability 2023 , 15 , 13091. [ Google Scholar ] [ CrossRef ]
• Sharma, A.; Khar, S. Design and development of a vegetable plug seedling transplanting mechanism for a semi-automatic transplanter. Sci. Hortic. 2024 , 326 , 112773. [ Google Scholar ] [ CrossRef ]
• Ji, D.; Liu, L.; Zeng, F.; Zhang, G.; Liu, Y.; Diao, H.; Tian, S.; Zhao, Z. Design and Experimental Study of a Traction Double-Row Automatic Transplanter for Solanum lycopersicum Seedlings. Horticulturae 2024 , 10 , 692. [ Google Scholar ] [ CrossRef ]
• Ji, D.; Tian, S.; Wu, H.; Zhao, B.; Gong, Y.; Ma, J.; Zhou, M.; Liu, W. Design and experimental verification of an automatic transplant device for a self-propelled flower transplanter. J. Braz. Soc. Mech. Sci. Eng. 2023 , 45 , 420. [ Google Scholar ] [ CrossRef ]
• Markumningsih, S.; Hwang, S.-J.; Kim, J.-H.; Jang, M.-K.; Nam, J.-S. Stress Simulation on Cam-Type Transplanting Device of Semiautomatic Vegetable Transplanter. Agriculture 2023 , 13 , 2230. [ Google Scholar ] [ CrossRef ]
• Miah, M.S.; Rahman, M.M.; Hoque, M.A.; Ibrahim, S.M.; Sultan, M.; Shamshiri, R.R.; Ucgul, M.; Hasan, M.; Barna, T.N. Design and evaluation of a power tiller vegetable seedling transplanter with dibbler and furrow type. Heliyon 2023 , 9 , e17827. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Liu, W.; Tian, S.; Wang, Q.; Jiang, H. Key technologies of plug tray seedling transplanters in protected agriculture: A review. Agriculture 2023 , 13 , 1488. [ Google Scholar ] [ CrossRef ]
• Paudel, B.; Basak, J.K.; Jeon, S.W.; Lee, G.H.; Deb, N.C.; Karki, S.; Kim, H.T. Working speed optimisation of the fully automated vegetable seedling transplanter. J. Agric. Eng. 2024 , 55 , 1569. [ Google Scholar ] [ CrossRef ]
• Sharma, A.; Khar, S.; Chaudhary, D.; Thakur, P. Study of biometric attributes of plug type tomato seedlings pertinent to transplanter design. Indian J. Ecol. 2023 , 50 , 503–507. [ Google Scholar ]
• Habineza, E.; Ali, M.; Reza, M.N.; Woo, J.-K.; Chung, S.-O.; Hou, Y. Vegetable transplanters and kinematic analysis of major mechanisms: A review. Korean J. Agric. Sci. 2023 , 50 , 113–129. [ Google Scholar ] [ CrossRef ]
• Elwakeel, A.E.; Mazrou, Y.S.; Eissa, A.S.; Okasha, A.M.; Elmetwalli, A.H.; Makhlouf, A.H.; Metwally, K.A.; Mahmoud, W.A.; Elsayed, S. Design and Validation of a Variable-Rate Control Metering Mechanism and Smart Monitoring System for a High-Precision Sugarcane Transplanter. Agriculture 2023 , 13 , 2218. [ Google Scholar ] [ CrossRef ]
• Li, M.; Xiao, L.; Ma, X.; Yang, F.; Jin, X.; Ji, J. Vision-based a seedling selective planting control system for vegetable transplanter. Agriculture 2022 , 12 , 2064. [ Google Scholar ] [ CrossRef ]
• Stubbs, S.; Colton, J. The Design of a Mechanized Onion Transplanter for Bangladesh with Functional Testing. Agriculture 2022 , 12 , 1790. [ Google Scholar ] [ CrossRef ]
• Chen, K.; Yuan, Y.; Zhao, B.; Jin, X.; Lin, Y.; Zheng, Y. Finite element modal analysis and experiment of rice transplanter chassis. Int. J. Agric. Biol. Eng. 2022 , 15 , 91–100. [ Google Scholar ] [ CrossRef ]
• Paradkar, V.; Raheman, H.; Rahul, K. Development of a metering mechanism with serial robotic arm for handling paper pot seedlings in a vegetable transplanter. Artif. Intell. Agric. 2021 , 5 , 52–63. [ Google Scholar ] [ CrossRef ]
• Li, P.; Yun, Z.; Gao, K.; Si, L.; Du, X. Design and Test of a Force Feedback Seedling Pick-Up Gripper for an Automatic Transplanter. Agriculture 2022 , 12 , 1889. [ Google Scholar ] [ CrossRef ]
• Yang, Q.; Zhang, R.; Jia, C.; Li, Z.; Zhu, M.; Addy, M. Study of dynamic hole-forming performance of a cup-hanging planter on a high-speed seedling transplanter. Front. Mech. Eng. 2022 , 8 , 896881. [ Google Scholar ] [ CrossRef ]
• Zhang, M.; Jiang, Z.; Jiang, L.; Wu, C.; Yang, Y. Design and Test of the Seedling Cavitation and Lodging Monitoring System for the Rape Blanket Seedling Transplanter. Agriculture 2022 , 12 , 1397. [ Google Scholar ] [ CrossRef ]
• Khadatkar, A.; Mathur, S. Design and development of an automatic vegetable transplanter using a novel rotating finger device with push-type mechanism for plug seedlings. Int. J. Veg. Sci. 2022 , 28 , 121–131. [ Google Scholar ] [ CrossRef ]
• Yang, J.; Zhou, M.; Yin, D.; Yin, J. Design and Development of Rice Pot-Seedling Transplanting Machinery Based on a Non-Circular Gear Mechanism. Appl. Sci. 2024 , 14 , 1027. [ Google Scholar ] [ CrossRef ]
• Zhao, X.; Hou, Z.; Zhang, J.; Yu, H.; Hao, J.; Liu, Y. Study on the Hole-Forming Performance and Opening of Mulching Film for a Dibble-Type Transplanting Device. Agriculture 2024 , 14 , 494. [ Google Scholar ] [ CrossRef ]
• Haili, Z.; Wei, Y.; Gaohong, Y.; Bin, W.; Bingliang, Y. Optimization Design and Experiments of Ditching Multi-bar Seedling Planting Mechanism. Trans. Chin. Soc. Agric. Mach. 2023 , 54 , 79–86. [ Google Scholar ]
• Lei, W.; Liang, S.; Yadan, X.; Gaohong, Y.; Gervais, N.L.; Jiahui, H. Multi-pose Motion Synthesis of Three-arm Gear Train Planting Mechanism Based on Genetic Algorithm. Trans. Chin. Soc. Agric. Mach. 2022 , 53 , 70–77. [ Google Scholar ]
• Gaohong, Y.; Zhenpiao, L.; Lehui, X.; Peng, Z.; Chuanyu, W. Optimization design and test of large spacing planetary gear train for vegetable pot-seedling planting mechanism. Trans. Chin. Soc. Agric. Mach. 2015 , 46 , 38–44. [ Google Scholar ]
• Gaohong, Y.; Ye, J.; Shushu, C.; Bingliang, Y.; Jinbo, G.; Xiong, Z. Design and test of clipping-plug type transplanting mechanism of rice plug-seedling. Trans. Chin. Soc. Agric. Mach. 2019 , 50 , 100–108. [ Google Scholar ]
• Hu, S.; Hu, M.; Yan, W.; Zhang, W. Design and experiment of an integrated automatic transplanting mechanism for picking and planting pepper hole tray seedlings. Agriculture 2022 , 12 , 557. [ Google Scholar ] [ CrossRef ]
• Shi, L.; Zhao, W.; Sun, W. Parameter calibration of soil particles contact model of farmland soil in northwest arid region based on discrete element method. Trans. Chin. Soc. Agric. Eng. 2017 , 33 , 181–187. [ Google Scholar ]
• Daqing, Y.; Nuoyi, Z.; Maile, Z.; Yuchao, Y.; Size, Y.; Jinwu, W. Optimal Design and Experiment of High Speed Duckbill Planting Mechanism with Variable Catch-seedling Attitude. Trans. Chin. Soc. Agric. Mach. 2020 , 51 , 66–72. [ Google Scholar ]
• Ji, J.; Yang, L.; Jin, X.; Gao, S.; Pang, J.; Wang, J. Design and parameter optimization of planetary gear-train slip type pot seedling planting mechanism. Trans. CSAE 2018 , 34 , 83–92. [ Google Scholar ]
• Long, Z.; Bo, Y.; Xin, J. Structure Design of Duck Mouth Transplanting End Actuator Based on TRIZ Theory. Agric. Eng. 2020 , 10 , 76–80. [ Google Scholar ]

Click here to enlarge figure

Share and Cite

Shi, J.; Hu, J.; Li, J.; Liu, W.; Yue, R.; Zhang, T.; Yao, M. Design and Experiment of Planting Mechanism of Automatic Transplanter for Densely Planted Vegetables. Agriculture 2024 , 14 , 1357. https://doi.org/10.3390/agriculture14081357

Shi J, Hu J, Li J, Liu W, Yue R, Zhang T, Yao M. Design and Experiment of Planting Mechanism of Automatic Transplanter for Densely Planted Vegetables. Agriculture . 2024; 14(8):1357. https://doi.org/10.3390/agriculture14081357

Shi, Jiawei, Jianping Hu, Jing Li, Wei Liu, Rencai Yue, Tengfei Zhang, and Mengjiao Yao. 2024. "Design and Experiment of Planting Mechanism of Automatic Transplanter for Densely Planted Vegetables" Agriculture 14, no. 8: 1357. https://doi.org/10.3390/agriculture14081357

Article Metrics

Further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Airbnb's ongoing battle with hotels for travelers has taken a noticeable turn

• This post originally appeared in the Insider Today newsletter.
• You can sign up for Business Insider's daily newsletter here .

Hello there! The iPhone 16 isn't even out yet, but we're already hearing rumors about the iPhone 17 lineup. Apple is reportedly working on an "Air" model that has gotten the Ozempic treatment and slimmed down considerably .

In today's big story, Airbnb said its struggles are tied to consumers spending less, but it's also because they're opting for hotels instead .

What's on deck:

• Markets: Inside a massive hedge fund's bid to keep a pipeline of strong talent .
• Tech: An excerpt from Nate Silver's new book examining prolific risk-takers .
• Business: Elon Musk and Donald Trump sat down for a rambling conversation on X .

But first, let's just book a hotel.

If this was forwarded to you, sign up here.

The big story

Plenty of vacancy.

Airbnb took a noticeable step back in its ongoing battle with hotels.

The short-term rental giant's stock is down almost 13% after lowering its third-quarter revenue projections and acknowledging it's seeing signs of slowing demand from US guests during last week's earnings call.

Execs quickly pointed to an economy that's seen consumers pull back on their spending habits. But Airbnb's struggles are also a product of travelers opting for hotels instead , writes Business Insider's Dan Latu.

That's quite a change from just a few years ago when Airbnb was booming as people looked to escape cities for extra space and fresh air. But the intervening years haven't always been kind to Airbnb.

A laundry list of chores and rules coupled with high cleaning fees sullied the experience for some . Bait-and-switch scams were another concern . But even if you were willing to look past all that, you might not even be able to book an Airbnb as an increasing number of cities have banned the platform .

The Airbnb-to-hotel conversion has been taking place for a while. But lowering its projections for a third quarter — a time when consumers are known to travel — felt like a turning point.

Airbnb isn't alone in facing a bit of a reckoning.

Apps that relied on low prices to grow fast and build loyalty have been forced to adjust their business model. The death of the so-called millennial lifestyle subsidy — for which Airbnb was a key player — resulted in consumers rethinking things.

So yes, customers have pulled back on spending in a way that has impacted Airbnb. But it's also a result of how Airbnb runs its business as opposed to just wider economic forces.

Still, weakening consumer spending is a lot easier to point to than acknowledging the competitor might be eating your lunch. A tough economy can be useful for businesses in that way, like how companies were able to raise prices on the premise of inflation going up.

And if you can't beat them, join them.

Airbnb signaled it might look to its rival for some inspiration. The company's chief business officer told Bloomberg it's looking at offering luxury services for guests in a bid to win hotel converts back.

Top headlines

• Trump claims he faced 'unacceptable intrusion' over FBI Mar-a-Lago raid in notice to sue the DOJ for $100 million .
• A US Navy carrier strike group equipped with F-35 stealth fighters is rushing to the Middle East along with a submarine packed with Tomahawk missiles .
• There's a 40% chance the US economy is already in a recession, according to a new indicator .
• The US housing market is on the verge of hitting a record $50 trillion valuation as prices keep rising .
• Felicis Ventures leads $10 million round in AI startup MemGPT at a $70 million valuation, sources say.

3 things in markets

• Bank of America isn't worried about all the spending on AI. The bank said concerns about the heavy investment into the tech are "premature." Traditionally, tech expenditures are front-loaded and don't always rely on creating new revenue streams , BofA analysts wrote.
• How a $48 billion hedge fund tried to flip the script on Wall Street's war for talent. Viking Global has gotten more aggressive when it comes to hiring analysts in a bid to keep its pipeline strong and promote from within . Viking's PMs are also partially compensated based on their analysts' career development.
• A simmering conflict in the Middle East has oil prices on the rise. WTI crude oil and Brent crude prices spiked ahead of a potential attack by Iran against Israel .

3 things in tech

• Investors <3 crazy tech founders. In his new book, "On the Edge: The Art of Risking Everything," Nate Silver writes that some founders have an unusual advantage: a chip on their shoulder. Coming from a background that's made them feel left out, excluded, or estranged can make them extremely competitive — a trait VCs are willing to bet on.
• Don't worry about upgrading to Apple's new iPhone 16. Bloomberg reported the new iPhone won't be much different from the iPhone 15 (which wasn't much different from the iPhone 14). The new models will have Apple Intelligence, but with some early reviews saying it's not worth the hype , some customers may be happy sticking with their older cell phones.
• AI has a big marketing problem. Advertisers love talking about AI. The problem? Consumers don't like hearing about it . Experts say that in order to effectively advertise for artificial intelligence, firms must focus on enhancing human creativity rather than replacing it.

3 things in business

• Canceling pesky subscriptions could get easier. Trying to get out of a gym membership or cancel a subscription can be a painstaking endeavor, but the White House just announced a proposal to make it a little easier . The new rule would make it as easy to cancel a subscription as it was to sign up for one.
• Three cofounders of the Alexander brothers' brokerage firm are jumping ship. The owners are leaving the firm, Official, after recent rape accusations against its star brokers Tal and Oren Alexander . Their departures follow unfruitful negotiations to remove the brothers from the business by having them relinquish their ownership stakes.
• Musk gave Trump 1 million listeners. Trump gave his greatest hits. Elon Musk's live-streamed conversation with former President Donald Trump on X began with technical difficulties — reminiscent of when Musk interviewed Florida Gov. Ron DeSantis on the site last year. Over the course of Monday's conversation, the pair discussed Trump's assassination attempt, a potential role for Musk in a second Trump administration, and more .

In other news

• An M&A expert says creator-economy companies need to look beyond influencers to grow .
• She grew up in Ukraine, studied accounting, and joined the army. But since Russia invaded, she did a hard career pivot — twice .
• Disney World is entering its villain era .
• I've been to over 200 high-end golf courses around the globe. Here are eight mistakes I always see first-timers make .
• Consulting has a Gen Z problem .
• Six things you could be doing wrong if you're struggling to get a job .

What's happening today

• The Home Depot and other companies are reporting earnings.
• "Made by Google" event to launch Google's latest devices, including the Pixel smartphone.

The Insider Today team: Dan DeFrancesco, deputy editor and anchor, in New York. Jordan Parker Erb, editor, in New York. Hallam Bullock, senior editor, in London.

Watch: The role of the CMO has grown more complex over the past five years says DoorDash CMO Kofi Amoo-Gottfried

• Main content