Genome mining reveals abiotic stress resistance genes in plant genomes acquired from microbes via HGT

Abstract

Colonization by beneficial microbes can enhance plant tolerance to abiotic stresses. However, there are still many unknown fields regarding the beneficial plant-microbe interactions. In this study, we have assessed the amount or impact of horizontal gene transfer (HGT)-derived genes in plants that have potentials to confer abiotic stress resistance. We have identified a total of 235 gene entries in fourteen high-quality plant genomes belonging to phyla Chlorophyta and Streptophyta that confer resistance against a wide range of abiotic pressures acquired from microbes through independent HGTs. These genes encode proteins contributed to toxic metal resistance (e.g., ChrA, CopA, CorA), osmotic and drought stress resistance (e.g., Na⁺/proline symporter, potassium/proton antiporter), acid resistance (e.g., PcxA, ArcA, YhdG), heat and cold stress resistance (e.g., DnaJ, Hsp20, CspA), oxidative stress resistance (e.g., GST, PoxA, glutaredoxin), DNA damage resistance (e.g., Rad25, Rad51, UvrD), and organic pollutant resistance (e.g., CytP450, laccase, CbbY). Phylogenetic analyses have supported the HGT inferences as the plant lineages are all clustering closely with distant microbial lineages. Deep-learning-based protein structure prediction and analyses, in combination with expression assessment based on codon adaption index (CAI) further corroborated the functionality and expressivity of the HGT genes in plant genomes. A case-study applying fold comparison and molecular dynamics (MD) of the HGT-driven CytP450 gave a more detailed illustration on the resemblance and evolutionary linkage between the plant recipient and microbial donor sequences. Together, the microbe-originated HGT genes identified in plant genomes and their participation in abiotic pressures resistance indicate a more profound impact of HGT on the adaptive evolution of plants.

Introduction

Plants, the most profoundly distributed lineage on earth, possess the remarkable abilities to convert solar energy into bio-matter and produce oxygen while consume the greenhouse gas carbon dioxide. At the same time, plants have to face the frequent offense of inhospitable abiotic stresses caused by the constant fluctuations of global climate since plants cannot “run away” from them. The challenging conditions like drought, radiation, nutrient deficiencies, temperature extremes (cold and heat), metal ion toxicity, salinity and organic pollution are typical abiotic stress factors that plants frequently encounter (Zhang et al., 2022). These abiotic factors have drastically influenced plant distribution and growth status globally (Nolan et al., 2018). It is long known that beneficial microbes are helping plants to overcome climate change and other adverse conditions through a variety of ways (Schützendübel and Polle, 2002; Haney et al., 2015; Murali et al., 2021). For example, microbes can increase the activity of plant antioxidant defense enzyme (Mastouri et al., 2012), release beneficial volatile organic compounds (VOCs) that trigger resistance response (Liu and Zhang, 2015), and modulate the expression of plant metal transporters against the accumulation of heavy metals (Zhou et al., 2020). However, many aspects of the plant-microbe interactions for mutual benefits are still in the “black box” and lack in-depth mechanism analysis.

Horizontal gene transfer (HGT), referring to the exchange of genetic material between distant lineages, is considered as an effective mechanism to spread evolutionary success in both prokaryotes and eukaryotes (Huang, 2013; Daubin and Szöllősi, 2016). Studies on HGT genes introduced into eukaryotic taxa have been expanding fast in the last twenty years as next-generation sequencing technology has made large-scale genome data available (Soucy et al., 2015). Recent researches have demonstrated that genes in plant genomes important for the function of specific plant activities, including the biosynthesis of plant structure components and hormones, plant defense, nitrogen recycling and stress resistance, can be obtained via the omnipresent HGT process (Yue et al., 2012). It is strongly proposed in a recent review that HGTs possess the potency to drastically influence plant evolutionary path by imparting selective advantages (Wickell and Li, 2020). Here are some examples illustrating the importance of microbe-derived HGTs in plants that confer adaptive benefits. Firstly, it is found that the ancestor of land plants acquired a phenylalanine ammonia lyase (PAL), the entry point of the phenylpropanoid pathway, through HGT during symbioses with soil bacteria and fungi (Emiliani et al., 2009); The second case is that Physcomitrella patens is suggested to acquire from bacteria acyl-activating enzyme 18 (AAE18)/flavin monooxygenase (YUC3) for auxin biosynthesis, guanine deaminase, allantoate amidohydrolase and ureidoglycolate amidohydrolase associated with purine degradation, as well as subtilases for protein degradation (Yue et al., 2012). Besides, Li et al. (2018) detected that the squalene-hopene cyclase (SHC) encoding genes in three plant clades (mosses, liverworts and lycophytes/ferns) are interspersing with bacterial SHCs, suggesting the occurrence of independent and repeated HGT events for enhanced adaption since the product of SHC, hopanoid, constitutes plant cell membrane important for maintaining cell integrity. Lastly, it is reported that the adaptive genes in Zygnematophyceae, GRAS and PYR/PYL/RCAR that enhance resistance to abiotic stresses (e.g., desiccation), were obtained through HGT from nearby soil bacteria (Cheng et al., 2019). However, many of these previous investigations focused mainly on individual plant lineage or limited HGT events on important economic aspects such as the valuable plant metabolism pathways. Despite these efforts, we still have little knowledge on the amount or impact of HGT-derived genes in plants that have potentials to confer resistance against a wide range of abiotic stresses (e.g., low pH, osmotic pressure, heat and cold stress) and hence promote the growth of plants in extreme conditions. It is still an open question that if there is microbe-originated HGT(s) that has facilitated plant to adapt such abiotic stresses. And if there are such HGT genes, what are their functions in combating abiotic stresses and to what extent are they distributed in the plant genomes?

Against such background, in this study, we have chosen fourteen high-quality plant genomes covering major plant taxa (Clade Viridiplantae) within Phyla Chlorophyta and Streptophyta, and tried to uncover and scrutinize putative microbe-originated HGT genes that are associated with abiotic stress resistance in these genomes employing a combination of methods previously showed to be highly effective in identifying HGT between distantly related species (Schönknecht et al., 2014), including BLASTP-based searches and gene tree reconciliation. To further testify the expression potential of the detected HGT genes, we supply codon adaption index calculation, deep-learning-driven protein structure prediction, fold alignment, and molecular dynamics to assess the expression potential and functionality of the abiotic stress resistance genes acquired by plants from microbial lineages. Our findings indicated that the HGT events in beneficial microbes putatively colonizing plant might be an important source of acquisitions of genes that help enhance plant tolerance against abiotic stresses. This study has also advanced our understanding on plant-microbe interactions and the adaptive evolution of plants to encounter adverse conditions.

Results

General information

Detailed information of the fourteen high-quality eukaryotic plant genomes covering major plant taxa (clade Viridiplantae) within phyla Chlorophyta and Streptophyta chosen for downstream analyses is provided in Table 1 . Putative microbe-originated HGT genes in these genomes employing the BLASTP-based IMGAP pipeline (Markowitz et al., 2010), followed by manual pickup of microbe-originated genes that are associated with abiotic stress resistance. These processes eventually produce a total of 235 gene entries, as listed in Table 2 and Supplementary Tables S1A and S1B at https://doi.org/10.6084/m9.figshare.20530083.v1. Detailed descriptions are provided in the following subsections.

Table 1.

High-quality eukaryotic plant genomes covering major plant taxa (clade Viridiplantae) within phyla Chlorophyta and Streptophyta chosen for microbe-originated horizontal gene transfer (HGT) detection in this study.

Genome name	Phylum	NCBI bioproject	NCBI genbank ID	IMG genome ID	Sequencing depth	Sequencing quality	Sequencing status	Genome size (bp)	Scaffold count	Gene count	GC %
Chlamydomonas reinhardtii CC-503 cw92 mt+	Chlorophyta	PRJNA12260	ABCN00000000	649410502	10x	Level 2: High-Quality Draft	Complete	120404952	1558	14546	63.87
Volvox carteri f. nagariensis 69-1b	Chlorophyta	PRJNA13109	ACJH00000000	2507525017	8x	Level 2: High-Quality Draft	Complete	137835555	1265	14542	55.97
Chlorella variabilis NC64A	Chlorophyta	PRJNA45853	ADIC00000000	2507525016	8.9x	Level 1: Standard Draft	Complete	46159515	414	9791	67.14
Arabidopsis thaliana Columbia	Streptophyta	PRJNA13190	GCA_000211275	639560100		Level 6: Finished	Complete	119707899	7	31370	36.06
Physcomitrella patens patens (Moss)	Streptophyta	PRJNA13064	ABEU00000000	649410504	8.63x	Level 2: High-Quality Draft	Complete	477845029	1992	35938	33.58
Glycine max cultivar Williams 82 (Soybean)	Streptophyta	PRJNA19861	ACUP00000000	2507525014	8.5x	Level 2: High-Quality Draft	Complete	973344380	1168	55787	34.75
Populus balsamifera trichocarpa (Balsam poplar)	Streptophyta	PRJNA10772	AARH00000000	649410505	7.5x	Level 2: High-Quality Draft	Complete	480871302	20644	40566	33.55
Zea mays mays cv. B73 (Maize)	Streptophyta	PRJNA10769	CM000786, CM000777, GK000031	2507525013	6x	Level 6: Finished	Complete	2065722704	11	106044	46.89
Arabidopsis lyrata lyrata MN47	Streptophyta	PRJNA41137	ADBK00000000	649410501	8x	Level 2: High-Quality Draft	Complete	206667935	695	32549	36.08
Vitis vinifera PN40024 (Grape Vines)	Streptophyta	PRJNA34679	CAAP00000000	649410508	12x	Level 2: High-Quality Draft	Complete	791124756	3535	24954	34.51
Ricinus communis Hale (Castor bean)	Streptophyta	PRJNA16585	AASG00000000	649410506	2.45x	Level 2: High-Quality Draft	Complete	350458699	25762	31894	33.84
Sorghum bicolor BTx623	Streptophyta	PRJNA13876	ABXC00000000	2507525011	4x	Level 2: High-Quality Draft	Complete	738540932	3304	29448	43.93
Solanum tuberosum DM 1-3 516 R44 (Potato)	Streptophyta	PRJNA63145	AEWC00000000	2507525040	114x	Level 2: High-Quality Draft	Complete	727424546	66254	24377	34.8
Fragaria vesca Hawaii 4 (Woodland strawberry)	Streptophyta	PRJNA60037	AEMH00000000	2507525039	49x	Level 2: High-Quality Draft	Complete	214219504	3254	7931	38.4

Table 2.

Microbe-originated HGT genes identified in tested plant genomes that associated with miscellaneous abiotic stress resistance.

Plant recipient	Recipient gene accession (IMG)	Gene function	Donor gene accession (IMG)	Microbial donor	Donor taxon
Metal resistance
Chlorella variabilis NC64A	2507982899	Magnesium-translocating P-type ATPase MgtA	646357403	Escherichia coli O111:H 11128	Bacteria; Proteobacteria
Chlamydomonas reinhardtii CC-503 cw92 mt+	649485460	Magnesium and cobalt transport protein	2508406796	Spizellomyces punctatus DAOM BR117	Eukaryota; Chytridiomycota
Chlamydomonas reinhardtii CC-503 cw92 mt+	649493513	Copper homeostasis protein CutC	2506487166	Emticicia oligotrophica GPTSA100-15	Bacteria; FCB group
Physcomitrella patens patens	649531133	Copper resistance protein CopC	648722241	Paenibacillus curdlanolyticus YK9	Bacteria; Firmicutes
Ricinus communis Hale	649585425	Chromate transporter ChrA	649771835	Methylovorus sp. MP688	Bacteria; Proteobacteria
Arabidopsis thaliana Columbia	639572405	Copper-transporting ATPase	645867480	Talaromyces marneffei ATCC 18224	Eukaryota; Ascomycota
Arabidopsis thaliana Columbia	639588060	P-type copper ATPase	645729519	Aspergillus clavatus NRRL 1	Eukaryota; Ascomycota
Volvox carteri f. nagariensis 69-1b	2507995811	Magnesium and cobalt transport protein	2508406796	Spizellomyces punctatus DAOM BR117	Eukaryota; Chytridiomycota
Vitis vinifera PN40024	649546378	Ca2+-transporting atpase	2509695751	Schizophyllum commune H4-8	Eukaryota; Opisthokonta; Fungi
Physcomitrella patens patens	649501412	Acid phosphatase	2509205796	Batrachochytrium dendrobatidis JEL423	Eukaryota; Chytridiomycota; Chytridiomycetes
Osmotic and drought stress resistance
Physcomitrella patens patens	649496502	Choline dehydrogenase BetA	2509380569	Ensifer aridi TW10	Bacteria; Proteobacteria; Deltaproteobacteria
Volvox carteri f. nagariensis 69-1b	2508002683	Chloride intracellular channel 6	639857298	Stigmatella aurantiaca DW4/3-1	Bacteria; Proteobacteria; Deltaproteobacteria
Volvox carteri f. nagariensis 69-1b	2507995432	Potassium/proton antiporter	645972328	Streptomyces sp. SPB78	Bacteria; Actinobacteria; Actinomycetia
Chlorella variabilis NC64A	2507981403	Sarcosine oxidase SoxA	646435156	Sphaerobacter thermophilus 4ac11, DSM 20745	Bacteria; Chloroflexi; Thermomicrobia
Chlorella variabilis NC64A	2507981739	Na+/proline symporter	2508394518	Allomyces macrogynus ATCC 38327 (fungus)	Eukaryota; Blastocladiomycota; Blastocladiomycetes
Chlorella variabilis NC64A	2507982753	Chloride channel protein ClcA	637771792	Synechococcus sp. CC9902	Bacteria; Cyanobacteria; Synechococcales
Arabidopsis thaliana Columbia	639580571	Na+/solute symporter	646259962	Oxalobacter formigenes HOxBLS	Bacteria; Proteobacteria; Betaproteobacteria
Arabidopsis thaliana Columbia	639566001	Na/K ATPase alpha 1 subunit, putative	645866880	Talaromyces marneffei ATCC 18224	Eukaryota; Ascomycota; Eurotiomycetes
Chlamydomonas reinhardtii CC-503 cw92 mt+	649489365	Sodium/calcium exchanger	649408713	Perkinsus marinus PmCV4CB5 2B3 D4	Eukaryota; Perkinsozoa; Perkinsida
Chlamydomonas reinhardtii CC-503 cw92 mt+	649487367	Sodium/proton antiporte, CPA1	637235733	Nostoc sp. PCC 7120	Bacteria; Cyanobacteria; Nostocales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649492246	Zinc-nutrition responsive transporter, ZIP family	649337865	Phaeodactylum tricornutum CCAP1055/1	Eukaryota; Bacillariophyta; Bacillariophyceae
Glycine max cultivar Williams 82	2507918029	Maltooligosyl trehalose hydrolase (EC 3.2.1.141)	651457131	Xanthomonas vesicatoria Maraite, ATCC 35937	Bacteria; Proteobacteria; Gammaproteobacteria
Arabidopsis thaliana Columbia	639571770	Trehalose 6-phosphate synthase (EC 2.4.1.15)	646433456	Thermanaerovibrio acidaminovorans Su883, DSM 6589	Bacteria; Synergistetes; Synergistia
Arabidopsis thaliana Columbia	639573047	Teichoic acid biosynthesis protein B	646070519	Staphylococcus aureus A9635	Bacteria; Firmicutes; Bacilli
Chlamydomonas reinhardtii CC-503 cw92 mt+	649484689	Cellulose synthase (UDP-forming)	637692951	Cupriavidus pinatubonensis JMP134	Bacteria; Proteobacteria; Betaproteobacteria
Chlorella variabilis NC64A	2507980628	dTDP-D-glucose 4,6-dehydratase RmlB	641205904	Acanthocystis turfacea Chlorella virus 1	Viruses; dsDNA viruses, no RNA stage; Megaviricetes
Physcomitrella patens patens	649498952	Capsular exopolysaccharide family	2505768802	Fischerella thermalis JSC-11	Bacteria; Cyanobacteria; Synechococcales
Chlorella variabilis NC64A	2507978560	Malate synthase (EC 2.3.3.9)	641353162	Sorangium cellulosum So ce 56	Bacteria; Proteobacteria; Deltaproteobacteria
Chlorella variabilis NC64A	2507978803	Malate synthase (EC 2.3.3.9)	641353162	Sorangium cellulosum So ce 56	Bacteria; Proteobacteria; Deltaproteobacteria
Chlorella variabilis NC64A	2507978611	Isocitrate lyase (EC 4.1.3.1)	643430771	Anoxybacillus flavithermus WK1, DSM 2641	Bacteria; Firmicutes; Bacilli
Physcomitrella patens patens	649526838	Transaldolase/Fructose-6-phosphate aldolase	647578161	Synechococcus sp. PCC 7335	Bacteria; Cyanobacteria; Synechococcales
Heat and cold stress
Glycine max cultivar Williams 82	2507922078	Stromal 70 kDa chaperone protein DnaK	643477038	Rippkaea orientalis PCC 8801	Bacteria; Cyanobacteria; Chroococcales
Chlorella variabilis NC64A	2507984527	DnaJ-class molecular chaperone with C-terminal Zn finger domain	645858879	Pyrenophora tritici-repentis Pt-1C-BFP	Eukaryota; Ascomycota; Dothideomycetes
Chlorella variabilis NC64A	2507978787	DNA-binding protein HU	644676974	Bacillus thuringiensis sv. monterrey BGSC 4AJ1	Bacteria; Firmicutes; Bacilli
Chlorella variabilis NC64A	2507982059	DnaJ-class molecular chaperone with C-terminal Zn finger domain	2507660422	Phytophthora ramorum Pr102, UCD Pr4	Eukaryota; Oomycota; Peronosporales
Chlorella variabilis NC64A	2507982566	Chaperonin GroEL (HSP60 family)	2509191772	Ectocarpus siliculosus Ec 32 (CCAP 1310/04)	Eukaryota; Phaeophyceae; Ectocarpales
Chlorella variabilis NC64A	2507983822	Small heat shock protein (HSP20) family	641146850	Marinobacter algicola DG893	Bacteria; Proteobacteria; Gammaproteobacteria
Chlorella variabilis NC64A	2507983994	DnaJ molecular chaperone homology domain	2507660422	Phytophthora ramorum Pr102, UCD Pr4	Eukaryota; Oomycota; Peronosporales
Chlorella variabilis NC64A	2507984993	Heat shock protein 70	2509695036	Schizophyllum commune H4-8	Eukaryota; Opisthokonta; Fungi
Volvox carteri f. nagariensis 69-1b	2507993038	DnaJ-class molecular chaperone with C-terminal Zn finger domain	642750380	Hydrogenobaculum sp. Y04AAS1	Bacteria; Aquificae; Aquificae
Fragaria vesca Hawaii 4	2508472453	Heat shock protein 70KD	638270255	Guillardia theta	Eukaryota; Cryptophyceae; Pyrenomonadales
Arabidopsis thaliana Columbia	639560148	Heat shock protein. Metallo peptidase. MEROPS family M48B	639002844	Nitrococcus mobilis Nb-231	Bacteria; Proteobacteria; Gammaproteobacteria
Chlamydomonas reinhardtii CC-503 cw92 mt+	649482610	Heat shock factor binding protein 1	639628046	Trypanosoma brucei brucei 927/4 GUTat10.1	Eukaryota; Euglenozoa; Kinetoplastea
Ricinus communis Hale	649563104	Chaperone clpb, heat shock protein	637065254	Escherichia coli O157:H7 EDL933 (EHEC)	Bacteria; Proteobacteria; Gammaproteobacteria
Ricinus communis Hale	649583866	Molecular chaperone	2509190845	Ectocarpus siliculosus Ec 32 (CCAP 1310/04)	Eukaryota; Phaeophyceae; Ectocarpales
Ricinus communis Hale	649584950	Cold-shock DNA-binding protein CspA	642599004	Paraburkholderia phymatum STM815	Bacteria; Proteobacteria; Betaproteobacteria
Volvox carteri f. nagariensis 69-1b	2507995359	Chitinase	650773356	Burkholderia gladioli BSR3	Bacteria; Proteobacteria; Betaproteobacteria
Chlamydomonas reinhardtii CC-503 cw92 mt+	649491057	Chitinase family 18 (EC:3.2.1.14)	637922967	Saccharophagus degradans 2-40	Bacteria; Proteobacteria; Gammaproteobacteria
Chlamydomonas reinhardtii CC-503 cw92 mt+	649492930	Chitinase	2508410311	Spizellomyces punctatus DAOM BR117	Eukaryota; Chytridiomycota; Chytridiomycetes
Ricinus communis Hale	649570195	Ribulose 1,5-bisphosphate carboxylase large subunit (EC 4.1.1.39)	647109286	Raphidiopsis brookii D9	Bacteria; Cyanobacteria; Nostocales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649488017	Mitochondrial carbonic anhydrase, beta type	648186377	Gloeothece verrucosa PCC 7822	Bacteria; Cyanobacteria; Chroococcales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649410718	Photosystem II protein VI	648390068	Limnospira indica PCC 8005	Bacteria; Cyanobacteria; Oscillatoriales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649410720	Photosystem II cytochrome b6/f complex subunit V	637799911	Synechococcus elongatus PCC 7942	Bacteria; Cyanobacteria; Synechococcales
Oxidative stress resistance
Volvox carteri f. nagariensis 69-1b	2507998527	Peroxin-13	645816130	Coprinopsis cinerea okayama7#130	Eukaryota; Basidiomycota; Agaricomycetes
Volvox carteri f. nagariensis 69-1b	2507998720	Haem peroxidase	2506751490	Thiothrix nivea JP2, DSM 5205	Bacteria; Proteobacteria; Gammaproteobacteria
Chlorella variabilis NC64A	2507978793	Glutathione S-transferase	2504722272	Pseudomonas fluorescens HK44	Bacteria; Proteobacteria; Gammaproteobacteria
Chlorella variabilis NC64A	2507980056	Peroxiredoxin Q/BCP	645811546	Coprinopsis cinerea okayama7#130	Eukaryota; Basidiomycota; Agaricomycetes
Physcomitrella patens patens	649527966	Cysteine synthase A	640878576	Parvibaculum lavamentivorans DS-1	Bacteria; Proteobacteria; Alphaproteobacteria
Chlorella variabilis NC64A	2507979706	Glutaredoxin	2509193249	Ectocarpus siliculosus Ec 32 (CCAP 1310/04)	Eukaryota; Phaeophyceae; Ectocarpales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649493089	Glutaredoxin, CGFS type	2508453811	Saprolegnia parasitica CBS 223.65	Eukaryota; Oomycota; Saprolegniales
Physcomitrella patens patens	649517225	L-glutamine synthetase (EC 6.3.1.2)	2505791185	Runella slithyformis LSU4, DSM 19594	Bacteria; FCB group; Bacteroidetes
Volvox carteri f. nagariensis 69-1b	2507994380	Homocysteine/selenocysteine methylase (S-methylmethionine-dependent)	651485911	Rubrivivax benzoatilyticus JA2	Bacteria; Proteobacteria; Betaproteobacteria
Zea mays mays cv. B73	2507869489	Glutamate synthase (NADPH/NADH)	649510728	Physcomitrella patens patens	Eukaryota; Streptophyta; Bryopsida
Glycine max cultivar Williams 82	2507946040	Glutamate synthase (NADH) large subunit (EC 1.4.1.14)	638941273	Vibrio sp. MED222	Bacteria; Proteobacteria; Gammaproteobacteria
Volvox carteri f. nagariensis 69-1b	2507995895	Folylpolyglutamate synthase	2508319981	Micromonas pusilla NOUM17, RCC 299	Eukaryota; Chlorophyta; Mamiellophyceae
Arabidopsis thaliana Columbia	649593200	Glutamate synthase (ferredoxin) (EC 1.4.7.1)	637875655	Synechococcus sp. JA-2-3B’a(2-13)	Bacteria; Cyanobacteria; Synechococcales
Arabidopsis thaliana Columbia	649604117	Glutamate synthase (ferredoxin) (EC 1.4.7.1)	640085243	Prochlorococcus marinus NATL1A	Bacteria; Cyanobacteria; Synechococcales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649485827	Thiol-disulfide isomerase and thioredoxins	2508407653	Spizellomyces punctatus DAOM BR117	Eukaryota; Chytridiomycota; Chytridiomycetes
Populus balsamifera trichocarpa	649557489	Cytochrome-c oxidase	645666500	Talaromyces marneffei ATCC 18224	Eukaryota; Ascomycota; Eurotiomycetes
Ricinus communis Hale	649585270	Electron transfer flavoprotein beta subunit	646832083	Methylotenera versatilis 301, JCM 17579	Bacteria; Proteobacteria; Betaproteobacteria
Volvox carteri f. nagariensis 69-1b	2507995502	Ubiquinol-cytochrome c reductase cytochrome b subunit	640951028	Rickettsia canadensis McKiel	Bacteria; Proteobacteria; Alphaproteobacteria
Volvox carteri f. nagariensis 69-1b	2508002619	Menaquinol-cytochrome c reductase iron-sulfur subunit precursor	639801009	Paenarthrobacter aurescens TC1	Bacteria; Actinobacteria; Actinomycetia
Physcomitrella patens patens	649529911	c-type cytochrome synthesis protein Ycf5	641534169	Microcystis aeruginosa NIES-843	Bacteria; Cyanobacteria; Chroococcales
Arabidopsis thaliana Columbia	639588215	Cytochrome c oxidase assembly protein	639887067	Synechococcus sp. BL107	Bacteria; Cyanobacteria; Synechococcales
Chlorella variabilis NC64A	2507979399	Phytoene dehydrogenase	641347970	Sorangium cellulosum So ce 56	Bacteria; Proteobacteria; Deltaproteobacteria
Arabidopsis thaliana Columbia	639581518	Isocitrate dehydrogenase, NADP-dependent	637833532	Hahella chejuensis KCTC 2396	Bacteria; Proteobacteria; Gammaproteobacteria
Chlorella variabilis NC64A	2507984961	Thiamine biosynthesis protein ThiC	2509194402	Ectocarpus siliculosus Ec 32 (CCAP 1310/04)	Eukaryota; Phaeophyceae; Ectocarpales
Volvox carteri f. nagariensis 69-1b	2508000640	Thiamine pyrophosphate enzyme	651583914	Sphingomonas sp. S17	Bacteria; Proteobacteria; Alphaproteobacteria
Arabidopsis thaliana Columbia	639576678	Thiamine pyrophosphate protein	646441210	Thermobaculum terrenum YNP1, ATCC BAA-798	Bacteria; Chloroflexi; Thermobaculum
Chlorella variabilis NC64A	2507979588	Thiamine monophosphate synthase	2502685596	Pontibacter actiniarum DSM 19842	Bacteria; FCB group; Bacteroidetes
Chlorella variabilis NC64A	2507980977	Selenocysteine lyase	646501278	Conexibacter woesei ID131577, DSM 14684	Bacteria; Actinobacteria; Thermoleophilia
Chlorella variabilis NC64A	2507981240	Selenocysteine lyase	2508443779	Saprolegnia parasitica CBS 223.65	Eukaryota; Oomycota; Saprolegniales
Volvox carteri f. nagariensis 69-1b	2507997145	Selenocysteine lyase	2509158209	Entamoeba histolytica HM-1:IMSS	Eukaryota; Evosea; Mastigamoebida
Chlamydomonas reinhardtii CC-503 cw92 mt+	649484656	Selenophosphate synthase	2507499783	Crocosphaera subtropica BH63E	Bacteria; Terrabacteria group; Cyanobacteria
Chlorella variabilis NC64A	2507981440	Cobalamin synthesis protein cobW	640895065	Roseiflexus castenholzii HLO8, DSM 13941	Bacteria; Chloroflexi; Chloroflexia
Chlamydomonas reinhardtii CC-503 cw92 mt+	649491021	Cobalamin synthesis protein cobW	640596696	Roseiflexus sp. RS-1	Bacteria; Chloroflexi; Chloroflexia
Physcomitrella patens patens	649500991	Cobalamin synthesis protein cobW	649849043	Pantoea sp. At-9b	Bacteria; Proteobacteria; Gammaproteobacteria
Acid resistance
Sorghum bicolor BTx623	2507715021	Spermidine/putrescine transporter	2509391205	Rhizobium leguminosarum bv. trifolii WSM597	Bacteria; Proteobacteria;
Volvox carteri f. nagariensis 69-1b	2508000101	Arginine deiminase ArcA (EC 3.5.3.6)	639619926	Dictyostelium discoideum AX4	Eukaryota; Evosea; Eumycetozoa
Chlamydomonas reinhardtii CC-503 cw92 mt+	649410730	Proton extrusion protein PcxA	640014237	Lyngbya sp. PCC 8106	Bacteria; Cyanobacteria; Oscillatoriales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649494337	Arginine deiminase ArcA (EC 3.5.3.6)	649376570	Naegleria gruberi NEG-M (Amoeba)	Eukaryota; Heterolobosea; Vahlkampfiidae
Chlorella variabilis NC64A	2507979843	Carbamate kinase (EC 2.7.2.2)	643475399	Rippkaea orientalis PCC 8801	Bacteria; Cyanobacteria; Chroococcales
Chlamydomonas reinhardtii CC-503 cw92 mt+	649494190	Polyamine transporter 3	649064649	Nannizzia gypsea CBS 118893	Eukaryota; Ascomycota; Eurotiomycetes
Physcomitrella patens patens	649499876	Basic amino acid/polyamine antiporter, APA family/solute carrier family 7 (L-type amino acid transporter), member 9	639684312	Candidatus Solibacter usitatus Ellin6076	Bacteria; Acidobacteria; Acidobacteriia
Physcomitrella patens patens	649531109	Argininosuccinate lyase	651569762	Paenibacillus sp. HGF7	Bacteria; Firmicutes; Bacilli
Ricinus communis Hale	649573998	Spermidine/putrescine transport system	2508732231	Microvirga lupini Lut6	Bacteria; Proteobacteria;
Ricinus communis Hale	649590952	Argininosuccinate synthase (EC 6.3.4.5)	640876950	Parvibaculum lavamentivorans DS-1	Bacteria; Proteobacteria; Alphaproteobacteria

HGT genes for metal resistance

The accumulation of toxic metals in contaminated soils stands for a serious environmental challenge, which negatively affects plant health and growth. As a possible solution, it has been reported that metal-resistant bacteria/fungi inhabiting plant-related sphere have the ability to promote plant growth and tolerance upon the attack of toxic metal stress (Mishra et al., 2017; Zhou et al., 2020), forming plant-microbe interactions which might also provide an opportunity for plants to acquire metal-resistant genes originated from nearby microbial taxa. In accordance, we identified eleven HGT gene entries for metal resistance in the tested plant genomes ( Table 2 ), most of which are metal ion transport proteins (e.g., ChrA, CopA, CorA) that facilitate the efflux of cytosolic excessed toxic metal ions. Other HGT genes related to metal resistance might include phosphatase ( Table 2 ), which might be secreted to increase soil phosphorus availability and immobilize toxic metal ions by forming complex compounds with them (Bechtaoui et al., 2021).

To confirm the HGT origin of the mentioned plant metal resistance genes, we selected and built the phylogenetic trees with three representative genes as queries: chromate transporters (ChrA) of Ricinus communis, copper homeostasis protein (CutC) of Chlamydomonas reinhardtii, and magnesium and cobalt transport protein (CorA) of Volvox carteri. As expected, all these suggested HGT genes in plant genomes are clustering with homologues of cross-phylum microbial taxa (putative HGT donors). To be specific, it is illustrated in the well-supported phylogenetic tree constructed with not-exclusive top BLASTP hit entries that: ChrA of Ricinus communis (eudicots) and several green algae (e.g., Chlorella spp., Chlamydomonas spp., Coccomyxa spp.) bear significant resemblance to the homologues of vast Proteobacteria and Firmicutes lineages (e.g., Methylophilus spp., Pseudomonas spp., and Bacillales) ( Figure 1A ); CutC homologues of green algae (e.g., Chlamydomonas spp., Volvox spp., Klebsormidium nitens) are completely nested within in bacterial taxa (e.g., Verrucomicrobiales, Bacteroidales, Rhizobiaceae), a pattern indicative of cross-kingdom HGT ( Figure 1B ); besides, CorA homologues of green algae (e.g., Chlamydomonas spp., Scenedesmus spp., Volvox spp.) are closely clustering with fungal lineages (e.g., Colletotrichum spp., Spizellomyces spp., Kwoniella spp.) ( Figure 1C ). Lastly, phosphatase homologues of green algae (e.g., Chlamydomonas spp., Ostreobium spp., Micromonas spp.) are clustering adjacent to bacterial taxa (e.g., Sphingomonas spp., Zoogloea spp., Pseudomonadales), indicating strong HGT signals ( Figure 1D ).

Figure 1.

Phylogenetic tree of metal resistance proteins in plant genomes originated from microbial taxa. The phylogeny construction was conducted based on gene-translating protein sequences using PhyML program with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values in red showing on respective branches). Branches leading to plant taxa are marked by green color, whereas those belonging to microbial lineages are in black: (A) chromate transporters (ChrA); (B) copper homeostasis protein (CutC); (C) magnesium and cobalt transport protein (CorA). (D) phosphatase.

Osmotic and drought stress resistance

Salinization and drought of arable land stand for major challenges to agriculture production worldwide. Salt accumulation and water scarcity would cause osmotic and water-deficit stress that restrict plant growth by affecting nutrient uptake and cell physiology (Bashir et al., 2014; Hanin et al., 2016). Acquisitions of osmotic and drought stress resistance genes from nearby microbes to enhance self-resistance might be an alleviating approach for plants. Accordingly, we found ten gene entries that putatively confer osmotic stress resistance in five tested plant genomes ( Table 2 ): eight ion/solute transporter (e.g., sodium/potassium ATPase, Na⁺/proline symporter, chloride channel protein and potassium/proton antiporter) and two organic osmolyte biosynthesis enzymes (choline dehydrogenase BetA and sarcosine oxidase SoxA). These genes are correspondent to two mechanisms of osmotic stress resistance: maintaining cellular ion homeostasis and recruitment of compatible solutes (e.g., proline, betaine, sarcosine) to stabilize proteins. These compatible solutes can also protect against drought, heat or cold stresses (Bashir et al., 2014; Hanin et al., 2016). Other putative HGT and osmotic/drought resistance genes detected in plants, as listed in Table 2 , may include: trehalose 6-phosphate synthase and trehalose hydrolase, and dTDP-D-glucose 4,6-dehydratase (RmlB), involving in the biosynthesis of protective sugar; cellulose synthase, significant for enhanced cell wall (Behr et al., 2015). Besides, we have found in the plant genomes HGT-driven malate synthase and isocitrate lyase of the glyoxylate cycle, which were reported to be highly expressed under drought conditions (Todaka et al., 2015; Kim et al., 2020); Lastly, transaldolase, an upstream enzyme controlling erythritol (protectant) production, which was found up-regulated upon exposure to short-term drought stress (Liu et al., 2015) and hyper-osmotic stress (Iwata et al., 2015; Yang et al., 2015).

To further verify the occurrence of HGT on the plant stress resistance genes mentioned above, we chose and built the phylogenetic trees of three representative genes: sodium/proton antiporter (CPA1) from Chlamydomonas reinhardtii, Na⁺/solute symporter from Arabidopsis thaliana, chloride channel protein (ClcA) and dTDP-D-glucose 4,6-dehydratase (RmlB) from Chlorella variabilis ( Figure 2 ). As observed from the well-supported phylogenetic tree of CPA1, homologues from Chlamydomonas spp. are tightly clustering with CPA1 from fungi (e.g., Basidiobolus, Rhizophagus, Actinomortierella), and remotely with CPA1 from cyanobacteria and actinobacteria ( Figure 2A ). Likewise, the well-supported phylogenetic tree of Na⁺/solute symporter placed green algae/plant homologues (e.g., Scenedesmus, Coccomyxa, and Klebsormidium) in the vicinity of those of fungi (e.g., Batrachochytrium, Podila, Mortierella) ( Figure 2B ). Similarly, in the well-supported phylogenetic tree of ClcA ( Figure 2C ), homologues from Chlorella spp. (green algae) clustering with several homologues of other green algae (e.g., Chlamydomonas, Micractinium), are embedded in branches derived from archaeal taxa of phylum Euryarchaeota (Methanobacteriales, Methanomassiliicoccales) and high G+C Gram-positive bacterial taxa (e.g., Pseudonocardia spp., Mumia spp. and Microbacterium spp.), suggesting the occurrence of cross-phylum HGT. Finally, the phylogenetic tree of RmlB indicates that Chlorella virus Acanthocystis turfacea has facilitated the transitions of RmlB encoding genes ( Figure 2D ), in consistence with previous studies (Parakkottil et al., 2010).

Figure 2.

Phylogenetic tree of osmotic and drought stress resistance proteins in plant genomes originated from microbial taxa. The phylogeny construction was conducted based on gene-translating protein sequences using PhyML program with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values in red showing on respective branches). Branches leading to plant taxa are marked by green color, whereas those belonging to microbial lineages are in black. (A) sodium/proton antiporter (CPA1); (B) Na⁺/solute symporter; (C) chloride channel protein (ClcA); (D) dTDP-D-glucose 4,6-dehydratase (RmlB).

HGT genes in response to heat and cold stress

Large fluctuations in environmental temperature caused by inconstant global climate also render severe abiotic impacts (heat and cold stress) on plants. To cope with such conditions, gene inventions and expansions of heat/cold shock factor gene families into plant genomes are constantly happening during adaptive evolution (Wang et al., 2018; Li et al., 2021). In this study, we have also detected in tested plant genomes a wide range (26 gene entries) of microbial-originated HGT gene families related to heat/cold shock response ( Table 2 ), which mainly function as transcription factors and molecular chaperones regulated by the former, working corporately to maintain cellular protein homeostasis (Andrási et al., 2021). In addition, it is reported that chitinase is also involved in plants’ cold and osmotic stress responses (Cao et al., 2019), and correspondingly, HGT gene entries encoding chitinases are also detected ( Table 2 ). In addition, we have discovered HGT genes encoding ribulose 1,5-bisphosphate carboxylase large subunit (Rubisco), carbonic anhydrase (CA), and photosystem II proteins putatively acquired from Bacteria ( Table 2 ). It is possible that plants might depend on generating more photosynthesis-related protein copies as a remedy strategy to cope with the impaired CO₂-fixing efficiency resulted from malfunction of heat-liable Rubisco under heat-stress conditions (Nouri et al., 2015).

As before, we chose and performed phylogenetic analyses on several such genes to illustrate the likely evolutionary path. The first gene query is heat-shock protein DnaJ, which is reported to protect Rubisco activity under heat stress (Wang et al., 2015). As expected, we found that DnaJ proteins from a group of green algae (e.g., Chlamydomonas, Tetrabaena, Edaphochlamys, Volvox) and eudicots (Erigeron canadensis) ( Figure 3A , marked with green color) are placed just next to homologues from bacterial lineages (e.g., Phycisphaerae, Anaerolineae, Pseudanabaenaceae). The second tested gene is heat shock protein Hsp20 ( Figure 3B ). Likewise, a group of green algae (e.g., Chlorella, Micractinium, Scenedesmus) is located in the well-supported tree of Hsp20 next to various bacterial taxa (e.g., Marinobacter spp., Marinimicrobia, Nitrospirae). The final case present the phylogeny of typical cold-shock protein CspA that promotes the correct folding of RNA molecules (Rennella et al., 2017). We found that several CspA-like proteins from eudicots (e.g., Ricinus communis, Lupinus albus) are in close homology to CspA proteins from Betaproteobacteria lineages (e.g., Burkholderiaceae, Methylophilaceae) ( Figure 3C ). Together, these phylogenetic relations might be existing molecular evidence supporting the HGT origins of plant abiotic resistance genes.

Figure 3.

Phylogenetic tree of heat and cold stress as well as oxidative stress resistance proteins in plant genomes originated from microbial taxa. The phylogeny construction was conducted based on gene-translating protein sequences using PhyML program with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values in red showing on respective branches). Branches leading to plant taxa are marked by green color, whereas those belonging to microbial lineages are in black: (A) heat-shock protein DnaJ; (B) heat shock protein Hsp20; (C) cold-shock protein CspA; (D) peroxiredoxin; (E) heme peroxidase (PoxA); (F) glutathione S-transferase (GST).

HGT genes resisting oxidative stress

Oxidative stress resulting from the excessive generation of reactive oxygen species (ROS) that damage cellular biomolecules, including proteins, DNA, lipids, and carbohydrates, is the usual consequence of miscellaneous abiotic stresses in plants mentioned above (e.g., drought (Niu et al., 2021), osmotic pressure (Cai-Hong et al., 2005), toxic metal stress (Schützendübel and Polle, 2002)). As illustrated in our results ( Table 2 ), it seems a potential strategy for plants to cope with the omnipresent oxidative stresses by obtaining directly from nearby microorganisms relevant resistance genes belonging to the canonical antioxidant machinery (e.g., hame peroxidase, glutaredoxin, thioredoxin, peroxin, peroxiredoxin, glutathione-S-transferase, cysteine synthase and glutamate/glutamine synthase) (Gill and Tuteja, 2010; Hasanuzzaman et al., 2019). Furthermore, miscellaneous HGT genes encoding components of oxidative electron transfer chain such as cytochrome oxidase (an oxygen-reducing reducing proton-pump) were also discovered ( Table 2 ), which are suggested to be involved in oxidative stress and acid stress tolerance (de la Garza-García et al., 2021). In addition, also discovered in tested plant genomes were bacteria/fungi-originated genes encoding an isocitrate dehydrogenase (in Arabidopsis thaliana), which might supply reductant NADPH to defend against oxidative stress (Komatsu et al., 2014), a phytoene dehydrogenase (in Chlorella variabilis) that converts phytoene into antioxidant carotene, a homocysteine S-methyltransferase (in Volvox carteri) that generates antioxidant methionine, as well as genes related to biosynthesis of antioxidant vitamins thiamine and cobalamin in multiple plants (Rosado-Souza et al., 2020; Vásquez et al., 2022), as well as selenophosphate synthase and selenocysteine lyase putatively conferring oxidative stress protection (Costa et al., 2011) ( Table 2 ).

As verified through phylogenetic analysis, the tested HGT genes in plants are all faithfully clustering with distant microbial taxa (the HGT donors), showing strong HGT signs ( Figure 3 ). To be specific, in the well-supported phylogenetic tree of peroxiredoxin, homologues from the green algae (e.g., Chlorella, Auxenochlorella, Micractinium) and charophytic algae Klebsormidium are embedded in those from fungi (e.g., Phakopsora, Puccinia, Irpex), which indicated that the green algae might have gained peroxiredoxin genes through gene exchange from nearby fungi ( Figure 3D ); Similarly, as seen from the reconstructed phylogeny of heme peroxidase (PoxA), a group of green algae (e.g., Gonium, Volvox, Scenedesmus, Ostreobium) is placed in the vicinity of multiple bacterial lineages (e.g., Thiothrix, Rhodospirillaceae, Anaerolineales) ( Figure 3E ); Lastly, in the phylogeny of glutathione S-transferase (GST), homologues from the green algae (e.g., Chlorella, Rhizophagus, Coccomyxa) are enveloped by those from multiple bacterial lineages (e.g., Pseudomonas, Achromobacter, Burkholderiales), again, suggesting the cross-kingdom HGT of GST genes among such taxa ( Figure 3F ).

HGT genes contributed to pH homeostasis

Environmental pH is considered as an important factor for plants since it influences soil physical, chemical properties and nutrient availability and thereby affects plant inner homeostasis and growth (Zhou et al., 2022). Interestingly, we have excavated seven microbe-derived gene entries linking with pH homeostasis maintenance and acid stress resistance in five tested plant genomes ( Table 2 ): two basic amino acid/spermidine/putrescine transporter and a proton extrusion protein (PcxA) encoding genes acquired from Bacteria, and genes within the arginine deiminase pathway (argininosuccinate synthase, argininosuccinate lyase, carbamate kinase and arginine deiminase) derived from Bacteria and Amoeba, which might conduct pH regulation by neutralization of proton with base substrates like ammonia, arginine and polyamine. Besides, the mentioned osmotic stress resistance ion/proton transporter above might also take part in adaptive responses to pH perturbation.

We further selected and constructed the phylogeny of two representative HGT genes, which encode arginine deiminase (ArcA) in Volvox carteri (chlorophyte algae) and basic amino acid/polyamine antiporter (YhdG) in Physcomitrella patens (mosses), to test the reliability of the HGT inference ( Figure 4 ). Consistency, the well-supported phylogenetic tree constructed with not-exclusive top hit entries showed that ArcA homologues of Volvox spp. are clustering with those of several green algae affiliations like Scenedesmus spp., Chlamydomonas spp., and Chlorella spp. etc. ( Figure 4A , marked with green color), which are wholly embedded in homologous sequences from protist (e.g., Carpediemonas, Fonticula, Plasmodiophora and Naegleria etc.), the most likely cross-phylum HGT donor(s), and also, remotely associated with a cluster of Bacteria (e.g., Desulfobacterales). Likewise, the well-supported phylogenetic tree of YhdG showed that homologues of Physcomitrella, closely clustering with homologues of other mosses (e.g., Sphagnum spp., Ceratodon purpureus), liverworts (Marchantia spp.), and a green plant (Klebsormidium nitens), are nested within homologues of fungi (e.g., Mucor spp. and Aspergillus spp.) and bacteria (e.g., Acidobacteria, Aminicenantes and Solimonas), strongly indicating that cross-phylum HGTs might contribute to the spread of YhdG from microbes to plants ( Figure 4B ).

Figure 4.

Phylogenetic tree of acid resistance and DNA damage resistance proteins in plant genomes originated from microbial taxa. The phylogeny construction was conducted based on gene-translating protein sequences using PhyML program with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values in red showing on respective branches). Branches leading to plant taxa are marked by green color, whereas those belonging to microbial lineages are in black. (A) arginine deiminase (ArcA); (B) basic amino acid/polyamine antiporter (YhdG); (C) chloride channel protein (ClcA); (D) dTDP-D-glucose 4,6-dehydratase (RmlB).

HGT genes contribute to ultraviolet light damage resistance

The plant DNA can be damaged by various abiotic factors such as chemical mutagens or ultraviolet (UV) light radiation due to long-term sun exposure, which would induce oxidative damage and cross-links (DNA–protein or DNA–DNA). The accumulation of such damages in the plant would ultimately cause genomic instability and cell death (Tuteja et al., 2009). Apart from cellular DNA-damage response pathways encoded by plants themselves, our results suggested that many plants have also recruited plenty of genes with functions related to DNA repair/replication/recombination (68 gene entries, see Table S1A at https://doi.org/10.6084/m9.figshare.20530083.v1). For example, we have also found in the reconstructed phylogeny of UvrD, a DNA helicase involved in excision repair of UV-induced DNA damage, that green algae homologues (e.g., Micractinium, Chlorella, Raphidocelis) are placed in the vicinity of homologues from cross-phylum Streptomyces spp. (fungi), which indicated the occurrence of HGT event ( Figure 4C ); Likewise, we found that several green algae (e.g., Volvox, Chlorella, Chlamydomonas) contain Rad25, a DNA helicase required for DNA repair and RNA polymerase II transcription, which is in close sequence homology to that from fungi (e.g., Podila, Lactarius, Auriculariopsis) ( Figure 4D ).

HGT genes contribute to organic pollutant resistance

Organic soil pollution discharged by anthropogenic processes like herbicides overuse and chemical waste mismanagement can threaten the life of plants, and the association of endophytic or mycorrhizal microbes with host plants can reduce the harmful effect of organic pollutants to plant (Sun et al., 2015; Rajtor and Piotrowska-Seget, 2016). Interestingly, we found that plants have integrated various enzymes related to organic pollutant catabolism and several multidrug transport proteins into their genomes, such as cytochrome P450s, laccase, beta-lactamase, alkanesulfonate monooxygenase, haloperoxidase, most likely through HGT from nearby microbes (81 gene entries, see Table S1B at https://doi.org/10.6084/m9.figshare.20530083.v1). The most commonly-seen HGT enzyme of these, cytochrome P450 (CytP450, six gene entries), is a group of oxygenase enzymes that can catalyze the catabolism of a wide range of organic pollutants using oxygen as the electron acceptor, undergoing various detoxification reaction, including hydroxylation, N-, O-, S-dealkylation, sulfurization, epoxidation, deamination, desulfurization, dehalogenation, peroxidation, and N- oxide reduction (Lin et al., 2022). In addition to microbial CytP450, two laccases for removal of phenolic pollutants (Asadgol et al., 2014) were found in the genomes of Chlamydomonas reinhardtii and Arabidopsis thaliana, respectively, putatively acquired from fungi. While other organic degradation oxidases we found acquired by plants from microorganisms might target more specific organic compounds, such as alkanesulfonate, amine and beta-lactams (see Table S1B at https://doi.org/10.6084/m9.figshare.20530083.v1).

As before, the HGT possibility was tested on representative genes. In the well-supported phylogenetic tree using CytP450 from Chlamydomonas reinhardtii (XP_001698815) as query and constructed with not-exclusive top 100 BLASTP hit entries showed that CytP450 homologues of green algae (e.g., Chlamydomonadales, Trebouxiales, Chlorellales, Sphaeropleales), along with several homologues from eudicots (Pentapetalae), are placed in the neighborhood of homologues from cyanobacteria (e.g., Synechococcales, Nostocales, Pseudanabaenales), clearly indicating the occurrence of cross-kingdom HGT events that have facilitated the acquisitions and spread of organic pollutant catabolism enzyme CytP450 from lower-grade microbial lineages to various plant taxa ( Figure 5A ).

Figure 5.

A case-study of microbe-originated plant cytochrome P450 and its donor protein. (A) Phylogenetic tree of the organic pollutant catabolism protein cytochrome P450 constructed by PhyML program with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values shown by size of symbols); (B) structure perdition (deep-learning-driven) and comparison of representative recipient plant cytochrome P450 (left, Chlamydomonas reinhardtii: XP_001698815) and its putative microbial donor (right, Synechocystis sp. PCC: WP_028946589); (C) ligand-protein simulation diagram (5 ns classical molecular dynamics) of cytochrome P450 from Chlamydomonas reinhardtii (XP_001698815) showing key residues in proteins interacting with the hame prothetic group (ligand). Interactions that occur more than 30.0% of the simulation time in the selected trajectory (0.0 through 5.0 ns), are shown; (D) ligand-protein simulation diagram (5 ns classical molecular dynamics) of the cytochrome P450 from donor Synechocystis sp. PCC (WP_028946589) showing key residues in proteins interacting with the hame prothetic group (ligand) for comparison.

Assessment of the functionality of HGT genes

To assess if the above-mentioned abiotic resistance genes in plants are functioning normally in plants, we resorted to protein structure analysis methods to assess fold reasonability on one hand and codon adaption index (CAI) calculation for prediction on gene expression potential of the abiotic stress resistance genes on the other hand.

We first tested if proteins encoded by such HGT-originated abiotic stress resistance genes in plant genomes can fold themselves into reasonable three-dimensional (3D) structures conveying acknowledged molecular functions. With the aid of the excellent deep-learning-driven structure prediction programs, AlphaFold2 (Jumper et al., 2021) and RoseTTAFold (Baek et al., 2021), we have generated full-length structure predictions of proteins encoded by all detected microbe-originated HGT genes in plants related to abiotic stress resistance in this study. The outcoming models of both methods in this test are available at https://doi.org/10.6084/m9.figshare.20530026.v1. The TM-align-based comparison showed that the predicted models from both methods (AlphaFold2 and RoseTTAFold) agreed well with each other, giving an average pairwise TM-score of 0.80 (TM-score > 0.5 indicates similar fold), indicating that our protein structural models are very reliable. We further used such protein models as inputs of the deep-learning-driven structure-based protein functional annotation program DeepFRI (Gligorijević et al., 2021) for functional assessment. Results showed that about 68.5% molecular function annotations of these HGT-driven proteins have reliable confident scores above the DeepFRI significance cutoff score of 0.5 (see Table S1C at https://doi.org/10.6084/m9.figshare.20530083.v1), strongly indicating that most of the aforementioned HGT-originated abiotic stress resistance proteins can fold themselves into a correct, recognizable, and functional 3D topology. Furthermore, TM-align-based structural alignments of the 3D models of protein homologue from corresponding microbial donors built by the same deep-learning method (The outcoming models are available at https://doi.org/10.6084/m9.figshare.20530026.v1) to the plant HGT-driven abiotic stress resistance proteins also reveal similar fold with a relatively high average pairwise TM-score of 0.72 (see Table S1D at https://doi.org/10.6084/m9.figshare.20530083.v1). To give more impression, we provided here a case study of the above-mentioned CytP450 from Chlamydomonas reinhardtii ( Figure 5B ). Comparisons of the CytP450 protein structure from the plant representative with that from the predicted microbial donor representative (Synechocystis sp. PCC) reveal a pair of folds showing highly identical topology with pairwise TM-score of 0.75 and average root mean square deviations (RMSD) of 1.99 Angstrom (Å). Resemblance was further found as we compared ligand-protein interaction patterns of these two structures by conducting classical molecular dynamics for 5 ns ( Figures 5C, D ). Results showed that both proteins used conserved residues like positively charged Arg and aromatic Tyr, Phe for ligand interactions throughout the simulation, suggesting strong evolutionary connection and reaction feasibility. Similar fold resemblances were found in structures derived from other pairs of HGT genes (see Figure S1 at https://doi.org/10.6084/m9.figshare.21008905.v1).

Due to translational selection, genes with high expression level generally have a stronger codon bias than those expressed at lower levels (Hiraoka et al., 2009; Zhou et al., 2013). The codon adaption index (CAI) initially proposed by Sharp and Li (1987) can measure synonymous codon usage bias for a tested gene sequence by comparing the similarity between the synonymous codon usage of a gene and the synonymous codon frequency of validated highly expressed reference gene sets. CAI value ranges from 0 and 1, with higher value indicating higher synonymous codon usage similarity with the reference/higher expression level (Puigbò et al., 2008). The CAI method has been successfully used in various areas, such as the predictions on expression likelihood (Puigbò et al., 2007; Puigbò et al., 2008) and level (Li et al., 2019) of heterologous/HGT gene, and deduction of lifestyles based on genomic data (Willenbrock et al., 2006). In this study, the expression levels of all predicted HGT genes conferring abiotic stress resistance in the plant genomes were assessed utilizing the codon adaption index (CAI) as a numerical estimator. Results showed that many HGT-driven abiotic stress resistance genes in plant genomes possess relatively high CAI values ( Figure 6 ): averagely, metal resistance (0.683), osmotic and drought stress resistance (0.682), acid resistance (0.712), heat and cold stress resistance (0.697), UV light damage resistance (0.702), and organic pollutant resistance (0.681). The HGT genes related with pH homeostasis and DNA repair have relative higher CAI values, which might indicate that the pH turbulence and UV light radiation are the most commonly encountered abiotic stresses.

Figure 6.

Ranges of codon adaption index (CAI) values of different kinds of abiotic resistance genes with EF-Tu gene as a reference.

Together, it means that the strategy of acquisitions of abiotic stress resistance genes by plants from microbes to enhance adaption might be successful and significant as far as our results corroborate.

Discussion

In this study, we have discovered an abundant repertoire of putative microbial-originated genes in 14 tested plant genomes that confer resistance to a wide range of abiotic stresses commonly encountered by plants, with the putative microbial HGT donors spanning from viruses, bacteria, fungi to protists ( Table 2 and see Table S1 at https://doi.org/10.6084/m9.figshare.20530083.v1). We have also applied in-silico approaches to verify the accuracy of blast-based HGT inference (phylogenetic analyses), the functionality of the 3D protein fold encoded by such HGT genes (deep-learning-driven protein structure prediction, fold alignment, molecular dynamics etc.), and the expression potential of HGT genes (codon adaption index calculation).

The emergence and mechanism of HGT events (“Why does lateral transfer occur in so many species and how?”) is recommended as a still-pending and significant scientific question worthy of more in-depth ivestigations by the editorial of the journal Science in the article “So much more to know” (American Association for the Advancement of Science, 2005), and we prudently considered that the present research might have supplied some hints for this question. HGT can serve as a significant force that supplies species with genetic innovations and regulation flexibility that guarantees successful migration into novel niches. In our results, we found that the most commonly seen microbial HGT donor taxon for acquisition of resistance genes is Burkholderiaceae (n=10). Members of Burkholderiaceae are characterized as plant-associated, which usually constitute the core floral community (Massoni et al., 2021). Besides, we found that Chlorophyta taxa (e.g. Chlorella variabilis NC64A, n=42) generally have more HGT resistance genes than lower-grade Streptophyta taxa (e.g. Physcomitrella patens patens (moss), n=21), which outnumber those from high-grade land plant (e.g. Ricinus communis Hale (castor bean), n=14). Such richness of HGT genes associated with abiotic stress resistance especially those within the transition-stage lineages might have reflected a ring linking to the early adaption processes of plant to accomplish the colonization of land environment from aquatic. These HGT genes seem well-suited to combat abiotic factors more frequently encountered in land environments (e.g., desiccation, fluctuations of temperature, UV radiation), in consistence with previous studies (de Vries et al., 2016; Pennisi, 2019; Bowles et al., 2020). The mechanism leading to such abundant cross-lineage gene transfer into plant genomes might be explained by the “weak-link model” (Huang, 2013). Namely, foreign DNA fragments (e.g., from microbes) could be naturally imported into the plant receptors at the susceptible stages in plant lifecycle, such as the single-cell/preliminary stages (germline cells, embryos, etc.) or dedifferentiation/asexual reproduction stages. The microbe-plant HGT processes can be promoted by intimate or direct physical contact through symbiosis, parasitism, infection, or other associations, especially when the weakly-protected cells are exposed to the environment. Once getting into the cell, fixation of HGT genes to the recipient genome can be facilitated by the still undergoing organellar DNA fragment integrations into the nuclear chromosome in the developing cells, followed by vertical transfer of integrated HGT genes to offspring cells via mitosis (Huang, 2013).

Concerns regarding such cross-lineage HGT genes might be the authenticity of both HGT direction and identification during the inference and the functionality of such HGT genes in the plant genomes. It has been argued that some of such HGT genes in the eukaryotic genome might result from endosymbiotic origin or contamination (Koutsovoulos et al., 2016).

To alleviate these concerns, we have integrated into our study the “gold standard” verification for the detected HGT genes: that a particular gene has been gained through HGT from a different lineage would be phylogenetic incongruence, where an evolutionary tree for a specific protein family is distinct from the established organismal phylogeny (Schönknecht et al., 2014). For example, the well-supported phylogenetic tree of chromate transporters (ChrA) (Fig. 1a) shows the protein sequence from Ricinus communis and green algae embedded within bacterial sequences. The most parsimonious explanation for this phylogenetic distribution is HGT from bacteria. Regarding the direction of the gene acquisition (microbe to plant instead of the reverse), it is ensured through the HGT detection pipeline IMGAP v.5.0 that only selects HGT candidate genes showing the best hits outside the taxonomic lineage of the tested genome (i. e. from distant phylum, class, etc.), but with lower-scoring or no hits within the lineage of the tested genome. The precise method for identification of HGTs by the IMGAP pipeline has been expatiated by Markowitz et al. (2010).

Furthermore, our supplemented phylogenetic analyses have also supported the direction of the predicted HGTs. As for excluding the possibility of false positive induced by contamination, we included in our analyses only high-quality complete (or near-complete) genome assemblies of plants ( Table 1 ), and our phylogenetic analyses further indicate that the historical HGT events not only occurred once but also can be confidently identified in cases in related plant taxa. For example, in the well-supported phylogenetic tree of copper homeostasis protein (CutC), the HGT-driven homologues were detected in a range of different green algae (from Chlamydomonas on the top of the tree to Coccomyxa inserted in the middle of the tree), again intertwined with branches leading to bacteria taxa. This suggested that such HGT events should be very likely actual since the probability that the said green algae’s genome sequences happen to contaminate simultaneously the same kind of DNA fragments in lab is minor.

As for the functionality of the HGT genes in the plant genomes, we have resorted to multiple methods (e.g., structure analysis and codon adaption index calculation) to illustrate the functionality of the HGT genes related to plant abiotic stress resistance. In line with our results, previous studies have experimentally verified that many horizontally transferred genes in plants Orobanchaceae, Cuscuta and Rafflesia are being expressed (Xi et al., 2012; Yang et al., 2016; Vogel et al., 2018; Yang et al., 2019). What’s even more surprising is that twelve of the HGT genes in Alloteropsis semialata (grass) genome have a higher expression level than their native homologues, and in one case, the native copy was replaced by the foreign one (Dunning et al., 2019). In consistence, our CAI calculation results also supported relatively high expression levels across the microbe-originated abiotic stress resistance genes in plants, which emphasized their significant functions in supporting the growth of the plant host in the face of harsh abiotic conditions.

In summary, the microbe-originated HGT genes in plant genomes identified in our analyses and their participation in resistance against diverse abiotic pressures commonly encountered in land environments indicate a widespread and profound impact of HGT on the evolution of plants and other eukaryotes. Still, our analysis presented here is not exhaustive. We believe that the microbe-to-plant HGT cases discovered by far stand for only the tip of the profound evolutionary iceberg. As more and more genomic sequences of superior quality become available for wider lineage branches in the tree of plants, it is foreseeable that future research will provide us with a more exciting panorama illustrating both the extent and the evolutionary significance of HGT in plants thoroughly.

Materials and methods

Fourteen high-quality eukaryotic plant genomes covering major plant taxa (clade Viridiplantae) within phyla Chlorophyta and Streptophyta, as listed in Table 1 were chosen and downloaded from the public database (Genbank/IMG) for downstream analyses. Identification of horizontally transferred genes in the genomes of plant genomes was conducted through the Integrated Microbial Genomes Annotation Pipeline (IMGAP) v.5.0 (Markowitz et al., 2010), which defined genes in tested plant genomes as having been horizontally transferred from a distant lineage with the principle: genes that have the best BLASTP hits (highest bit scores) or >90% of the best hits found outside the taxonomic lineage of the tested genome (i.e., from distant phylum, class, etc.) and with lower-scoring hits or no hits within the lineage.

The phylogeny of different kinds of abiotic resistance gene was constructed based on gene-translating protein sequences using PhyML program (Guindon et al., 2010) at http://www.phylogeny.fr/simple_phylogeny.cgi with the Maximum Likelihood (ML) method and 1,000 bootstrap replicates (with bootstrap values in red showing on respective branches) and visualized with iTOL (Letunic and Bork, 2021) at https://itol.embl.de/. Sequences were aligned with MUSCLE (Edgar, 2004) and trimmed with Gblocks (Talavera and Castresana, 2007) prior to tree construction.

AlphaFold2 (Jumper et al., 2021) and RoseTTAFold (Baek et al., 2021) were applied to generate the full-length 3D structure of proteins encoded by targeted HGT genes in plant genomes and their suggested donors, following guidance from https://github.com/deepmind/alphafold and https://github.com/RosettaCommons/RoseTTAFold. Docking of ligands into respective protein structure was conducted with AutoDock Vina v.1.2.0 (Trott and Olson, 2010). Molecular dynamics (MD) simulations for the targeted protein-ligand complex were performed using the Desmond Molecular Dynamics System, version 3.6, (D. E. Shaw Research, New York, NY, 2008) with OPLS_2005 force field (with default parameters) and visualized by Visual Molecular Dynamics (VMD) software v.1.9.4 and LigPlot+ (Laskowski and Swindells, 2011). DeepFRI (Gligorijević et al., 2021) was further applied for reliable structure-based functional annotation verify recognizable functional folds in the proteins encoded by the HGT genes related to abiotic pressure resistance in plant genomes.

Codon adaption index (CAI) was used as a numerical estimator of gene expression level (Hiraoka et al., 2009; Zhou et al., 2013), and correspondingly, the webserver CAIcal (Puigbò et al., 2008) (http://genomes.urv.es/CAIcal/calc.php) was applied to calculate respective CAI values of HGT genes in plant genomes participating in resistance of diverse abiotic pressures with the EF-Tu gene (the elongation factors) as refence, which is supposed to be highly expressed across most organisms (Puigbò et al., 2008), as conducted previously (Li et al., 2019). Besides, the Origin Pro 2020 software (OriginLab, Northampton, MA, USA) was used for data analysis and figure creation.

Data availability statement

The data presented in the study are available in the Genbank and IMG repositories under accession numbers provided in Table 1 .

Author contributions

LL, and HY conceived and designed the research. LL, SP, ZW, TZ, HL, YX, JL and YL analyzed the data. LL wrote the manuscript. All authors contributed to the article and approved the submitted version.

Funding

This research was supported by the key project of Science and Technology of Hunan Branch of China National Tobacco Corporation (202104, XX2022-2024Aa01, HN2021KJ05, 20-22A02, HN2020KJ02), the key research and development program of Hunan Province (grants no. 2020WK2022, 2022SK2076), the National Natural Science Foundation of China (NSFC) (No. 41807332), Fundamental Research Funds for the Central Universities of Central South University (no. 2022ZZTS0420) and Hunan International Scientific and Technological Cooperation Base of Environmental Microbiome and Application (No. 2018WK4019).

Acknowledgments

We are grateful for resources from the High-Performance Computing Center of Central South University.

Conflict of interest

Author ZW is employed by Zhangjiajie Tobacco Company of Hunan Province. Author TZ is employed by Hunan Urban and Rural Environmental Construction Co., Ltd. Authors YX and JL are employed by Chenzhou Tobacco Company of Hunan Province.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The authors declare that this study received funding from Science and Technology of Hunan Branch of China National Tobacco Corporation. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary Material

The Supplementary Material for this article can be found online at: https://doi.org/10.6084/m9.figshare.20530083.v1