Introduction

Phytochrome Interacting Factors (PIFs) belong to the basic Helix-Loop-Helix (bHLH) family of transcription factors. The bHLH protein domain consists of two segments: the basic region, required for DNA binding, and the helix-loop-helix, responsible for hetero and homodimerization (Toledo-Ortiz et al., 2003). PIFs are also characterized by the presence of a protein domain that interacts with photoactivated phytochromes (Favero, 2020). This interaction promotes the degradation of PIFs in the light and is crucial for their role as major regulators of light-regulated biological processes (Leivar and Monte, 2014). In etiolated seedlings, which germinate and develop in subterranean darkness, PIFs are active and repress photomorphogenic features, such as chlorophyll biosynthesis, cotyledon expansion, and repression of hypocotyl elongation (Leivar et al., 2008; Shin et al., 2009).

Besides their well-known function in adjusting seedling development to light, PIFs, particularly PIF4, have over the past years also been implicated in the regulation of different biological processes such as immunity (Gangappa et al., 2017), morphological adaptations to high ambient temperatures (Koini et al., 2009), stomatal development (Casson et al., 2009), leaf senescence (Sakuraba et al., 2014), freezing tolerance (Lee and Thomashow, 2012), salt tolerance (Wang et al., 2025), anthocyanin biosynthesis (Liu et al., 2015), or fatty acid biosynthesis (Liao et al., 2025). PIF4 has thus emerged as a key integrator of multiple external and internal signals to optimize plant development (Choi and Oh, 2016; Lucyshyn and Wigge, 2009).

Several studies have described the molecular mechanisms that control PIF4 gene expression, protein levels, and activity (Leivar and Quail, 2011; Paik et al., 2017; Pham et al., 2018), but its posttranscriptional regulation has never been characterized. Here we show that alternative splicing, a posttranscriptional process generating multiple mRNAs from the same gene, produces two different PIF4 transcripts specifically in response to heat stress.

Temperature deviations from the optimal range significantly impact plant development and survival. Increases in temperature are classified as either high ambient temperature or excessively hot temperatures. High ambient temperature is typically 5-6ºC above the optimum temperature (22ºC for Arabidopsis thaliana), while excessively hot temperatures exceed this range (Li et al., 2018). These distinct temperature ranges activate independent signaling pathways, leading to different physiological outcomes. Warm temperatures induce thermomorphogenesis, which generally promotes growth and development in a PIF4-dependent manner (Quint et al., 2016). Conversely, excessively hot temperatures trigger stress-responsive pathways aimed at adjusting growth and physiology to mitigate the negative effects of heat (Kan et al., 2023). To date, the role of PIFs in temperature signaling has centered on thermomorphogenesis, with only a few studies having explored the role of PIF proteins in heat stress responses (Li et al., 2021; Yang et al., 2022). Intriguingly, our results reveal that heat stress induces photomorphogenic features in etiolated seedlings and that this developmental response is mediated by an alternative splice form of PIF4.

Results

PIF4 is alternatively spliced in response to heat stress

Our analysis of the exon and intron positioning in all 15 members of the XV subfamily of Arabidopsis thaliana bHLH transcription factors that includes PIFs (Toledo-Ortiz et al., 2003) showed that, with the sole exception of the less conserved member HFR1, the intron/exon distribution of the bHLH domain is maintained. Its 150-bp long sequence is distributed among three exons, following an invariable proportion: 22%, 44% and 34% (Figure 1A). In addition, the middle exon containing the largest section of the bHLH domain is always 66-bp long and surrounded by phase 0 introns, which preserve codon identities and therefore the reading frame (Figure 1A). Hence, alternative splicing of this exon will produce protein isoforms differing only in the bHLH region (Supplemental Figure 1). Given the genomic particularities of the bHLH middle exon in these genes, we investigated its splicing regulation by quantifying its PSI (Percent Spliced In; percentage of transcripts that include the exon) in publicly available Arabidopsis thaliana RNA-seq samples covering several environmental conditions and tissues at different developmental stages (Supplemental Table 1). This analysis revealed skipping of the exon exclusively in two genes: PIF4 and PIF6 (Figure 1B and Supplemental Table 1). In the case of PIF6, although exon skipping occurs in nearly all tissues and conditions tested, this alternative splicing event has only been shown to be functional in seeds and embryos (Penfield et al., 2010), where PIF6 is highly expressed (Supplemental Figure 2). Strikingly, we found that the PIF4 gene undergoes alternative splicing in this particular exon, exon 5, exclusively when plants are under heat stress (Figure 1B and Supplemental Figure 3). Given that the protein arising from this exon skipping event will lack a portion of the bHLH domain (Supplemental Figure 1), we expect heat-induced alternative splicing to reduce the amounts of functional PIF4.

Figure 1.

Heat stress induces photomorphogenesis in the dark

To study the impact of the exon skipping event in the bHLH domain of PIF4, we applied heat stress (37 ºC) to 3-day-old etiolated seedlings, a developmental stage at which PIFs are known to be functional in repressing photomorphogenesis (Leivar et al., 2008; Shin et al., 2009). First, we confirmed that the heat-induced PIF4 alternative splicing event occurs in etiolated seedlings as well (Figure 2A), also showing that it does not occur in response to light and is sustained along time (Figure 2A and 2B). Remarkably, heat treatment of dark-grown seedlings partially induces photomorphogenesis — cotyledons open and hypocotyl elongation is repressed (Figure 2C and 2D). These morphological changes, although less pronounced, are characteristic of seedlings lacking PIF activity, as is the case with etiolated seedlings transferred to light or dark-grown quadruple pif1pif3pif4pif5 (pifq) mutants (Figure 2B and Supplemental Figure 4) (Leivar et al., 2009, Leivar et al., 2008; Shin et al., 2009). Moreover, we quantified protochlorophyllide (Pchlide), the phototoxic chlorophyll precursor that, when overaccumulated in etiolated seedings, leads to photobleaching upon light exposure. This analysis revealed higher levels of Pchlide and increased photobleaching in wild-type (WT) etiolated seedlings exposed to heat for 24 hours prior to light exposure, a trend also partially phenocopying pifq mutants (Figure 2E, 2F and Supplemental Figure 5) (Leivar et al., 2009). These phenotypes are consistent with heat-induced alternative splicing reducing the amounts of functional PIF4. Analysis of PIF1, PIF3, PIF4 and PIF5 expression levels under 37ºC demonstrated that none of these genes were transcriptionally downregulated by heat stress (Figure 2G and Supplemental Figure 6), discarding a strong reduction in PIF transcript levels as the cause of the observed phenotypes. Heat-stressed etiolated seedlings are phenotypically more similar to higher order pif mutants than to pif4 single mutants (Supplemental Figure 7 and 8) (Leivar et al., 2008; Shin et al., 2009), suggesting that the PIF4-S splice form generated by exon skipping may act as a dominant negative rather than being merely inactive. Previous studies have reported that different PIF4 protein isoforms can exert dominant negative effects, inhibiting the activity of other PIFs, and that alternative splicing can produce dominant negative transcription factor isoforms (Gangappa et al., 2017; Kim et al., 2020; Nicolas et al., 2015; Seo et al., 2011). However, direct evidence is needed to confirm that the PIF4 short isoform functions as a dominant negative, reducing the activity of the long isoform and other PIFs. Nevertheless, because heat-induced phenotypes are milder than those observed in pifq seedlings (Figure 2C, 2E and 2F), a substantial fraction of PIFs likely remains functional at 37 ºC. This is consistent with heat-induced alternative splicing affecting only around 50% of transcribed PIF4 mRNAs (Figure 2A and 2B).

Figure 2.

Heat-induced photomorphogenesis depends on PIF4 alternative splicing

To confirm the implication of PIF4 alternative splicing in the physiological changes undergone by heat-stressed etiolated seedlings, we quantified the morphological and chlorophyll-related phenotypes of transgenic plants expressing predominantly the long PIF4 splice form in the pif4-101 mutant background, (PIF4p::PIF4-L; Figure 3A and Supplemental Figure 9). Importantly, both the heat-induced Pchlide accumulation and cotyledon opening were strongly reduced in these lines, while the repression of hypocotyl elongation was maintained (Figure 3B, 3C and Supplemental Figure 10). In agreement with the reduction of Pchlide in PIF4-L plants, a non-significant but correlating trend in their photobleaching phenotypes was also observed (Figure 3B). Importantly, PIF4-L.1 expresses the long isoform at levels similar to those of WT plants (Supplemental Figure 9A), ruling out the possibility that the suppression of heat-induced phenotypes (cotyledon opening and Pchlide accumulation) is due to elevated PIF4 expression levels. In addition, consistent with the comparable alternative splicing levels observed in heat-stressed WT and pif4 seedlings (Figure 3A and Supplemental Figure 9B), the skipped exon is located upstream of the pif4-101 mutation (Supplemental Figure 9C), and the phenotypes are also comparable (Figure 3). Similar results were obtained with the other commonly used pif4 mutant (pif4-2; Leivar et al., 2008), which harbors a similar insertion site (Supplemental Figure 9D). Overall, our results substantiate a role for PIF4 alternative splicing in controlling heat-induced developmental responses in etiolated seedlings. Next, we quantified the transcriptional response to heat stress in etiolated seedlings with different levels of the long PIF4 splice form (Figure 3A). Interestingly, we found a strong enrichment of heat-regulated genes among those reported as PIFq-regulated or PIFq-bound (Pfeiffer et al., 2014) (Figure 3D). Furthermore, heat-induced transcriptional changes in pif4 mutants, the genetic background of PIF4-L seedlings, were significantly attenuated in these transgenic lines, yet the response remained far from abolished (Figure 3E). This result could be explained by some heat-induced transcriptional changes being fully PIF-independent, as shown in Figure 3D, and others being PIF-dependent but unaffected due to the considerable fraction of PIFs still functional under heat stress. Either scenario would also explain the partial reversion of the heat-induced phenotypic responses observed in PIF4-L lines.

Figure 3.

To confirm the role of PIF4 alternative splicing in regulating heat-induced responses, we generated transgenic plants expressing the short isoform of PIF4 under the control of its endogenous promoter (PIF4p::PIF4-S) and evaluated their morphology in the dark under control temperature conditions. These transgenic lines (PIF4-S) showed higher PIF4 expression levels than the corresponding WT control (Figure 4A), and in all three lines, the short isoform was the predominantly expressed variant (Figure 4B). We then conducted a phenotypic analysis of seedlings grown in the dark for 3 and 4 days at 21ºC. Interestingly, our results showed that enhanced production of the short isoform consistently promoted cotyledon opening, while changes in the hypocotyl length were not always detectable (Figure 4C). Thus, the cotyledon phenotype of these plants resembles that of WT plants exposed to heat stress (Figure 2C and 2D), linking the production of this isoform with heat-induced morphological adaptations. Notably, cotyledon opening in these transgenic plants at 21ºC is less pronounced than in heat-stressed plants (37ºC), indicating that the production of this isoform is not the unique mechanism underlying heat-induced cotyledon opening.

Figure 4.

Discussion

Our study reveals, for the first time, that cotyledon opening is a developmental response of etiolated seedlings exposed to heat stress. Heat stress also exerts a repressive effect on hypocotyl elongation in etiolated seedlings, a phenomenon previously reported but not extensively studied (Hong and Vierling, 2000; Karayekov et al., 2013; Larkindale et al., 2005; Martín and Duque, 2022). Karayekov et al. linked this inhibitory effect on hypocotyls to altered functioning of light signaling components such as CONSTITUTIVE PHOTOMORPHOGENIC 1 (COP1), ELONGATED HYPOCOTYL5 (HY5) and circadian clock components. Notably, they also proposed a physiological rationale for this response: dark-germinated seedlings approaching the soil surface may become more susceptible to heat shock episodes due to their proximity to sunlight. They hypothesized that the molecular mechanisms triggered by these episodes would prime seedlings for imminent light exposure. Given that the first exposure to sunlight represents a critical phase for etiolated seedlings, requiring rapid adaptation to ensure survival, we agree that the early activation of photomorphogenesis-associated traits could be advantageous. Our finding that heat promotes cotyledon opening, another hallmark of photomorphogenesis (Arsovski et al., 2012), supports the hypothesis that, in etiolated seedlings, heat may function as a signal to induce photomorphogenesis. In addition, we show that heat enhances Pchlide accumulation in dark-grown seedlings. Although the conversion of Pchlide to chlorophyllide is light-dependent, Arabidopsis seedlings accumulate Pchlide in the dark to expedite this process upon light exposure (Sperling et al., 1997). However, the amount of Pchlide must be tightly balanced with the availability of the enzyme that catalyzes its conversion to prevent the generation of reactive oxygen species and subsequent cellular damage upon illumination (Mochizuki et al., 2010; Reinbothe et al., 1996). This raises the question of whether heat-induced Pchlide accumulation in etiolated seedlings is an adaptative mechanism to accelerate chlorophyll production and optimize the transition to autotrophic development, or whether it is a side effect of prematurely activating light signaling in the dark, an outcome that, as suggested by increased photobleaching, would negatively impact seedling survival.

Our data also indicate that the induction of these photomorphogenic traits in etiolated seedlings depends, at least in part, on a heat-specific regulatory event: the alternative splicing of PIF4. This represents a novel finding, as PIF4, despite being one of the most studied proteins in Arabidopsis, has been reported to be regulated only at the transcriptional and post-translational levels (Balcerowicz, 2020; Favero, 2020). Notably, our phenotypic analyses of seedlings with altered patterns of PIF4 alternative splicing (PIF4-L and PIF4-S) suggest that the heat-induced isoform plays a role specifically in controlling cotyledon-related phenotypes. This implies that the mechanisms reported by Karayekov to control hypocotyl elongation under heat stress may operate in parallel with the alternative isoform of PIF4. Because our transcriptomic experiment was conducted using whole seedlings, we were unable to assess organ-specific effects in detail. We hypothesize that this alternative splicing event may be organ-specific or, alternatively, that the protein encoded by the heat-induced PIF4 isoform may be preferentially active in cotyledons due to specific protein interactors and/or molecular targets. Further research into organ-specific dynamics is needed to elucidate why this alternative splicing event appears to predominantly affect cotyledon development. These findings would provide valuable insights into the organ-specific roles of PIF proteins, an emerging area of research (Dong et al., 2019; Sun et al., 2016; Zhang et al., 2021).

Our study focuses on the role of PIF4 under heat stress, a condition in which its function remains poorly understood. In fact, the few previous reports linking PIF function to heat stress have yielded contrasting results, with PIFs acting as either positive (Li et al., 2021) or negative (Yang et al., 2022) regulators of the response. Here, we significantly expand the current understanding of this transcription factor by demonstrating its involvement in heat stress and and further reinforcing its role as a key integrator of diverse environmental signals in the regulation of plant development (Paik et al., 2017). Importantly, modulating alternative splicing to alter isoform abundance is emerging as a promising strategy for developing stress-resilient plants (Alhabsi et al., 2025). In this context, investigating the role of the PIF4 alternative isoform in heat-stressed adult plants, along with molecular strategies that specifically target the splicing sites involved in its regulation, could reveal a novel molecular target and offer an alternative genetic approach to enhance plant stress tolerance.

Methods

Plant materials

The Arabidopsis thaliana pif1-1pif3-3pif4-2pif5-3 (pifq) and pif4-2 mutant was obtained from the Nottingham Arabidopsis Stock Centre (NASC). PIF4-L transgenic plants expressing the PIF4 coding region driven by the endogenous promoter, together with its respective pif4 mutant control (pif4-101; Lorrain et al., 2008) were kindly provided by U. Pedmale (Cold Spring Harbor Laboratory, USA). PIF4-S transgenic plants were generated by PCR amplification and cloning a 1227-bp fragment containing the coding sequence region of the short splice variant under the control of a 2505-bp fragment upstream the PIF4 start codon corresponding to the PIF4 promoter, in the eGFP-tagged version of the binary pBA002 vector using the XbaI/AatI restriction sites (5’-GACGTTTCTAGAATGGAACACCAAGGTTGGAG-3’and 5’-GTGACGTCCGAGTGGTCCAAACGAGAAC-3’). The pPIF4 promoter was insertved into the promoterless pBA002 via HindIII/XbaI restriction sites (5’-TGTGAAGCTTCCAAAGTAATAAAAGTTGCCACAAC-3’and 5’-GACGTTTCTAGAGTCAGATCTCTGGAGACATTTC-3’). The resulting constructs were introduced into Agrobacterium tumefaciens strain EHA105 and subsequently used for agroinfiltration-mediated transformation of Col-0 seedlings (Clough and Bent, 1998).

Phenotypical and photobleaching analysis

Sterile seeds were sown on MS medium containing 1X Murashige and Skoog (MS) salts (Duchefa Biochemie), 2.5 mM MES (pH 5.7), 0.5 mM myo-inositol, and 0.8% agar (w/v). After stratification for 4 days at 4 ºC in darkness, seeds were subjected to a 3-hour light pulse to induce germination and then transferred to continuous darkness for 69 hours at 22 ºC. Maintaining the absence of light, seedlings were then either kept at 22 ºC for control conditions or transferred to 37 ºC for heat stress. Hypocotyl and cotyledon measurements of at least 30 seedlings and two biological replicates were carried out using the National Institutes of Health ImageJ software as described before (Sentandreu et al., 2011). Pictures were taken before and after exposure to stress as indicated in each figure. Photobleaching experiments were adapted from previous studies (Leivar et al., 2009), with 3-day-old etiolated seedlings being grown under control conditions or at 37 ºC for 24 hours and then transferred to continuous white light for 3 additional days (100 μmol·m⁻²·s⁻¹). At this point, the percentage of seedlings that failed to become green were scored in three biological replicates.

Protochlorophyllide quantification

Approximately 30 sterile seeds, sown on MS medium and stratified for 4 days at 4 ºC in the dark, were subjected to a 3-hour light pulse before being transferred to continuous darkness for 69 hours at 22 ºC. Seedlings were then either kept under these conditions (control) or transferred to 37 ºC in the dark to induce heat stress. Whole seedlings were collected in the dark 24 hours later, flash-frozen in liquid N₂,and ground before extraction with 0.75 mL ice-cold 9:1 acetone:0.1 M NH4OH, as described previously (Terry and Kacprzak, 2019). The resulting mixture was vortexed for 1 minute and then centrifuged at 14,000 rpm at 4 ºC for 5 minutes. After supernatant recovery, the protochlorophyllide (Pchlide) content was determined as the peak value (635 nm) of the fluorescence emission spectrum between 600-700 nm, measured with a bandwidth of 5 nm after excitation at 440 nm and using a Synergy Neo2 microplate reader (Biotek). Pchlide data is shown as the average of Pchlide per seedling of four biological replicates.

Gene expression and PSI quantification from RNA extraction

Total RNA was extracted from Arabidopsis thaliana seedlings using the InnuPREP Plant RNA kit (Analytik Jena BioSolutions) and 1 µg treated with DNase I to remove genomic DNA. cDNA synthesis using the oligo dT primer and the enzyme SuperScript III reverse transcriptase (Invitrogen) was conducted in the presence of RNase Out (Invitrogen). The cDNA was then used to quantify either gene expression or exon skipping of PIF4’s fifth exon. In both cases, three biological replicates were analyzed for each condition and/or genotype tested. Gene expression was measured by Reverse Transcription-quantitative PCR (RT-qPCR) using a QuantStudioTM 7 Flex Real-Time PCR System 384-well format and the Absolute SYBR Green ROX Mix (Thermo Scientific) on 2.5 µL of cDNA (diluted 1:10) per 10 µL of reaction volume, containing 300 nM of each gene-specific primer (see below). The PP2A gene was used for normalization (Shin et al., 2007). Exon skipping of the fifth exon of PIF4 was quantified from RT-PCRs using primers spanning the two adjacent exons. These primer sequences were obtained from PastDB (Plant alternative splicing and transcription Data Base; www.pastdb.crg.eu; Martín et al., 2021). The resulting bands were quantified using the National Institutes of Health ImageJ software.

RNA sequencing

RNA was extracted from 3-day-old WT, pif4-101, PIF4-L.1 and PIF4-L.2 etiolated seedlings grown for 3 hours at 37 ºC or 22 ºC in complete darkness. Oligo dT, non-strand specific libraries from triplicate biological replicates were built and sequenced using NextSeq500 at the Gulbenkian Institute for Molecular Medicine (GIMM). An average of 15 million 75-nucleotide single-end reads were generated per sample. Raw sequencing data was submitted to the Sequence Read Archive (accession number GSE200247).

Gene expression quantification from RNA sequencing data

Quantification of Arabidopsis thaliana transcript expression from our RNA-seq experiment (GSE200247) and public sequencing data (Dataset S1) was performed using vast-tools v2.5.1 and v2.2.2 (Martín et al., 2021; Tapial et al., 2017), respectively. For each Arabidopsis transcript, this tool provides the number of mapped reads per million mapped reads divided by the number of uniquely mappable positions of the transcript (cRPKM; corrected-for-mappability reads per kbp of mappable sequence per million mapped reads) (Labbé et al., 2012). To identify genes differentially expressed between different temperatures, we used vast-tools compare_expr using the option -norm to perform a quantile normalization of cRPKM values between samples. Next, we filtered out the genes that were not expressed at cRPKM > 5 and had read counts > 50 across all the replicates of at least one of the samples compared. Finally, differentially-expressed genes were defined as those with a fold change of at least 2 between each of the individual replicates from each genotype. See https://github.com/vastgroup/vast-tools for details.

PSI quantification from RNA sequencing data

We employed vast-tools v2.2.2 to quantify alternative splicing from public sequencing data (Martín et al., 2021; Tapial et al., 2017). This tool quantifies exon skipping (ES), intron retention (IR) and alternative donor (ALTD) and acceptor (ALTA) site choices. For all these types of events, vast-tools estimates the Percent Spliced In (PSI) of the alternative sequence using only exon-exon (or exon-intron for IR) junction reads and provides information about the read coverage See https://github.com/vastgroup/vast-tools for details. Data shown in Figure 1 and Supplemental Table 1 indicate the PSI quantification of specific alternative splicing events in the subfamily XV of the bHLH transcription factors (see below) using a wide array of samples (Supplemental Table 1).

Data availability

RNA-seq data have been deposited in Gene Expression Omnibus (GEO) (GSE200247).

Acknowledgements

We thank U. Pedmale for kindly providing pif4-101 mutants and PIF4pro:PIF4-3xFlag transgenic lines, and V. Nunes for excellent plant care at the Gulbenkian Institute for Molecular Medicine (GIMM) Plant Facility. This work was funded by Fundação para a Ciência e a Tecnologia (FCT) through grants PTDC/BIA-FBT/31018/2017, PTDC/BIA-BID/30608/2017 and PTDC/ASP-PLA/2550/2021 as well as by the Spanish Ministry of Science and Innovation trough grant PID2021-125223NA-I00 (MCIN/AEI/10.13039/501100011033/FEDER). Funding from the research unit GREEN-it “Bioresources for Sustainability” (ID/04551/2025, UID/PRR/04551/2025) and the Generalitat de Catalunya (AGAUR, GRE2021, ref. SGR00873) is also acknowledged. G.M. was supported by an EMBO Long-Term Fellowship (ALTF 1576-2016), a Marie Skłodowska-Curie Individual Postdoctoral Fellowship (EU project 750469) and a Ramón y Cajal fellow from the Spanish Ministry of Science and Innovation (RYC2021-032539-I). T.L was supported by a Marie Skłodowska-Curie Individual Postdoctoral Fellowship (EU project 706274).

Additional information

Author contributions

M.N.-G., B.A., D.S., T.L. and G.M performed the experiments and analyzed the data. All authors discussed the results. G.M. conceived the project and designed research. G.M. and P.D. wrote the manuscript.

Funding

FCT - Fundação para a Ciência e a Tecnologia (PTDC/BIA-BID/30608/2017)

• Guiomar Martín

FCT - Fundação para a Ciência e a Tecnologia (PTDC/BIA-FBT/31018/2017)

• Paula Duque

FCT - Fundação para a Ciência e a Tecnologia (PTDC/ASP-PLA/2550/2021)

• Paula Duque

FCT - Fundação para a Ciência e a Tecnologia (ID/04551/2025,UID/PRR/04551/2025)

• Paula Duque

Spanish Ministry of Science and Innovation (PID2021-125223NA-I00)

• Guiomar Martín

EMBO (ALTF 1576-2016)

• Guiomar Martín

Spanish Ministry of Science and Innovation (RYC2021-032539-I)

• Guiomar Martín

MSCA-IF European Commission (750469)

• Guiomar Martín

MSCA-IF European Commission (706274)

• Tom Laloum

Additional files

Supplementary file. Supplemental Figures 1-10 and Supplemental Tables 1-2.

Supplemental Table 1. List of publicly available RNA sequencing samples used. Samples available at the Short Read Archive (SRA) were used to quantify the PSI (Percent Spliced In) and expression values shown in Figure 1, Supplemental Figure 2 and Supplemental Figure 3. Columns indicate our sample classification based on their growth conditions and tissue source, which are summarized in the sample description column, as well as the original SRA code number and name. For simplification PSI quantification was performed in samples originated from merging the biological replicates.