🧬 biology

Does climate shape plant immune gene repertoires? A phylogenetically controlled test across angiosperms

This study demonstrates that while climate and geography are not robust predictors of immune receptor gene content across angiosperms once phylogenetic relationships are accounted for, woody and tree growth forms are consistently associated with elevated counts of specific cell-surface receptor genes, particularly LysM-RLKs.

Original authors: Stanley SX Tan

Published 2026-07-14

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Stanley SX Tan

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine plants as the ultimate stay-at-home residents of the biological world. They can't run away from a hungry bug or a sneaky fungus, and they don't have a mobile immune system like ours that sends white blood cells on patrol. Instead, they rely on a massive, static security team made of genes: tiny molecular "guards" stationed on their skin and inside their cells, ready to sound the alarm if an invader shows up.

For a long time, scientists had a hunch about where these guards would be most needed. The theory was simple: The tropics are the danger zone. The idea was that plants living in warm, wet, low-latitude jungles face a constant barrage of pathogens, so they should have evolved huge arsenals of these immune genes. Meanwhile, plants in cold or dry places would have smaller, more relaxed security teams.

But when Stanley Tan, a researcher at Yale, decided to put this idea to the test across hundreds of flowering plants, the plot twist was unexpected.

The "Tropical Danger" Myth Gets Busted
Tan looked at the genetic blueprints of 584 different plant species and checked their immune gene counts against their local weather. He found that while there were some loose connections in the raw data, they were like faint whispers. The moment he accounted for the plants' family trees—because relatives tend to look alike and live in similar places—those whispers vanished completely.

In fact, the data actually hinted at the opposite of the original theory: plants in warmer climates seemed to have fewer of certain helper immune genes, not more. The paper concludes that once you factor in who is related to whom, climate and geography are not reliable crystal balls for predicting how many immune guards a plant has. The "warmer is more dangerous" rule doesn't hold up when you look at the whole family tree.

The Real Hero: Being a Tree
If the weather isn't the boss, what is? The answer turned out to be the plant's lifestyle, specifically whether it's a tree or a herb.

Think of an herb like a mayfly: it lives fast and dies young, often completing its life cycle in a single season. A tree, on the other hand, is like a centuries-old fortress. It has to stand its ground for decades or even centuries, facing the same pathogens year after year.

The study found that woody plants and trees consistently carry more of a specific type of immune gene called LysM-RLK compared to their herbaceous cousins. This wasn't just a fluke; the signal was strong enough to show up even after rigorous statistical checks. It's as if the long-lived trees decided, "We're going to be here for a long time, so we need a bigger, more specialized security force."

Why the Extra Guards? A Mystery
Here is where the story gets tricky. The paper explicitly rules out one obvious explanation: It's not because trees are just sitting in a "super-pathogen" zone. When the researcher checked global databases for fungal infections, trees didn't actually have more pathogen exposure than herbs. In some cases, herbs were exposed to more soil fungi. So, the extra genes aren't just a reaction to a dirtier neighborhood.

So, why do trees have more LysM-RLK genes? The paper suggests two possibilities, but it can't prove which one is right yet:

The Immunity Hypothesis: Trees need more guards to fight off the endless waves of attackers they face over their long lives.
The Friendship Hypothesis: These same genes are also used to make friends with helpful fungi (mycorrhizae) and bacteria. Maybe trees just need more "handshake" genes to maintain their underground alliances.

Because the current data mixes these two functions together, the paper can't tell us if the trees are buying more guns or more party invitations. It's a genuine mystery that future research needs to solve.

The Bottom Line
The paper's main takeaway is a bit of a reality check for plant ecologists: Don't assume the weather dictates the immune system. The strongest pattern found was that being a long-lived tree comes with a specific genetic upgrade in cell-surface receptors.

However, the authors are careful not to call this a "solved" problem. While the link between being a tree and having more LysM-RLK genes is robust and consistent, it's still a "modest" effect. It didn't pass the strictest possible statistical test when looking at all 39 different immune gene families at once. It's a solid clue, a strong hint, but not the final word. The paper suggests that the next big step is to figure out if those extra genes are for fighting enemies or making friends, a question that remains wide open.

Technical Summary: Does climate shape plant immune gene repertoires? A phylogenetically controlled test across angiosperms

Problem Statement
The number of immune receptor genes in flowering plants (angiosperms) varies by more than an order of magnitude, ranging from dozens to over a thousand Nucleotide-Binding Leucine-Rich Repeat receptors (NLRs) and Pattern Recognition Receptors (PRRs) per genome. A widely held macroecological expectation posits that plants in warmer, wetter, lower-latitude environments experience higher pathogen pressure and consequently maintain larger immune repertoires. However, empirical evidence for this gradient is mixed. A critical confounding factor is that both immune gene counts and climatic niches are phylogenetically inherited; closely related species tend to share similar climates and similar gene counts. Without explicit phylogenetic correction, observed correlations may reflect shared ancestry rather than adaptive responses to the environment. Furthermore, the role of plant functional traits, specifically the distinction between woody (tree/shrub) and herbaceous growth forms, in shaping immune investment remains under-explored at a macroecological scale.

Methodology
The study utilized a recently published, uniformly annotated catalogue of immune receptor genes across 808 angiosperm species. The analysis focused on 39 annotated immune receptor families, including six NLR families, eight top-level Receptor-Like Kinase (RLK) families, 21 LRR-RLK subclasses, and four Receptor-Like Protein (RLP) families.

Data Filtering: Special-lifestyle/habitat (SLH) species (e.g., aquatic, parasitic, carnivorous) were excluded from primary analyses to avoid convergent reductions in immune repertoires that could confound macroecological patterns. This resulted in datasets of 584 species for climate analysis and 176 species for trait analysis.
Predictors:
- Climate/Geography: 18 variables derived from WorldClim v2.1 (medians and standard deviations of 19 bioclimatic variables) and geographic coordinates (latitude, range).
- Functional Traits: Categorical traits (woodiness, growth form) from GIFT and continuous traits (leaf/stem properties) from the TRY database.
- Pathogen Exposure: Independent estimates of ambient fungal pathogen pressure were derived from GBIF occurrence records and GlobalFungi soil metabarcoding data.
Statistical Framework:
- Bivariate Screening: Spearman rank correlations (continuous) and Kruskal-Wallis tests (categorical) were performed across all receptor families and predictors, followed by joint Benjamini-Hochberg (FDR) correction.
- Phylogenetic Correction: Significant associations were re-evaluated using Phylogenetic Generalized Least Squares (PGLS) models with Pagel's $\lambda$ estimated by maximum likelihood. BUSCO completeness was included as a covariate to control for genome assembly quality.
- Multiple Testing Strategy: A two-tier correction approach was employed. A confirmatory test was conducted within 12 principal receptor families (including the pre-specified focal family, LysM-RLK), while an exploratory sweep covered all 39 families.
- Sensitivity Analyses: Models were tested against branch-length transformations, taxon subsampling, domestication status, genome size, and ancestral state reconstruction.

Key Results

Climate and Geography:
- Bivariate screening identified 56 significant correlations out of 702 tests. However, all effect sizes were modest ( $|\rho| \le 0.24$ ).
- Crucially, none of the climate-receptor associations survived phylogenetic correction (PGLS). For instance, the correlation between annual mean temperature and TNL genes ( $\rho = -0.237$ ) became non-significant ( $p = 0.51$ ) after accounting for phylogeny.
- The direction of the strongest raw correlations often contradicted the hypothesis: warmer climates were associated with fewer helper-NLR genes (TNL, NRG1, RNL), not more.
- A negative correlation between CNL genes and latitude was identified as a taxonomic artifact (Simpson's paradox), driven by the over-representation of monocots (which lack TNL genes) at lower latitudes.
Functional Traits (Growth Form):
- In contrast to climate, growth form showed a more consistent association with immune gene content.
- LysM-RLK: Woody and tree species carried significantly more LysM-RLK genes than non-woody and herbaceous species. This signal was robust across non-parametric tests (FDR $p = 0.004$ ) and PGLS models within the 12 principal families ( $p = 0.007$ ).
- Other Receptors: Malectin-RLK and several LRR-RLK subclasses also showed elevated counts in woody/tree species in non-parametric tests.
- Statistical Significance: While the LysM-RLK signal was robust to confounding variables (branch length, domestication, genome size), it did not survive the conservative joint FDR correction across all 39 families (FDR $p = 0.13$ ). The authors interpret this as a real but modest effect that is statistically significant as a pre-specified hypothesis but not at the genome-wide scale.
Pathogen Exposure:
- Analysis of ambient fungal pathogen exposure (via GBIF and GlobalFungi) revealed that woody species did not experience greater pathogen pressure than herbs; in some metrics, exposure was actually lower in woody species.
- Adding pathogen exposure metrics to the PGLS models did not alter the growth-form coefficient, indicating that the association between woodiness and LysM-RLK is not mediated by differential exposure to ambient fungal pathogens.

Significance and Claims
The paper concludes that, across angiosperms, climate and geography are not robust predictors of immune receptor gene content once shared ancestry is accounted for. The apparent macroecological gradients in immune gene numbers are largely artifacts of phylogenetic structure rather than direct adaptive responses to environmental pathogen pressure.

The most consistent ecological correlate identified is growth form. Woody and tree species tend to carry expanded repertoires of specific cell-surface receptors, particularly LysM-RLK and Malectin-RLK. However, the study explicitly states that the aggregate gene counts cannot resolve whether this expansion reflects increased investment in pathogen immunity or symbiotic signaling (mycorrhizal/rhizobial), as LysM receptors function in both pathways. The authors note that distinguishing these mechanisms requires subfamily-level annotation (LYK vs. LYR) or targeted sampling, which was beyond the scope of the current dataset.

The study serves as a cautionary example for macroecological genomics, demonstrating that without rigorous phylogenetic controls, correlations between genomic traits and environmental variables can be misleading. It highlights that while growth form is a clearer predictor than climate, the effect sizes are modest, and the mechanistic basis for the observed patterns remains an open question.

More like this