Introduction

Understanding the response of living organisms to spaceflight is important for improved environmental engineering and the selection of organisms in biological life support systems. RNA plays a central role in organizing the response of living organisms to their environment. Therefore, omics analysis of RNA (transcriptomics) allows for a holistic assessment of the environmental responses at the molecular level. In the context of transcriptomics of spaceflight responses, the function of identified differentially expressed transcripts could be individually investigated, or they can be collectively used in gene enrichment analyses to identify the affected biological processes.

Many studies have used transcriptomics platforms to evaluate the response of model organism Arabidopsis thaliana to spaceflight. Results from these studies have suggested an array of affected processes, with prominent reports of defense response, cell wall remodeling, photosynthesis, and oxidative stress among them (Manzano et al. 2022; Baba et al. 2022). While many studies have used statistical methods to identify the impacted processes, statistical justifications have only sometimes been provided. RNA extraction from various plant samples, including whole seedlings, roots, hypocotyls, leaves, and cell lines, have been used in spaceflight transcriptomics investigations (Manzano et al. 2022). However, to our knowledge, only a study by Paul et al. has evaluated the transcriptomic response of the root-tip to spaceflight (Paul et al. 2017).

In their study, Paul et al. examined the response in the root-tips of two Arabidopsis ecotypes, Columbia-0 (C) and Wassilewskija (W), as well as that of the PhyD mutant of Col-0 (P) (Paul et al. 2017). The study includes three samples per each of these three genotypes in light and dark conditions on the International Space Station (ISS) and three samples per genotype in light and dark conditions in the ISS Environment Simulator (ISSES) Chamber at Kennedy Space Center. In the parent study, Paul et al. identified differentially expressed genes (DEGs) within each genotype, followed by an investigation of overlapping DEGs across genotypes. This approach has the potential to investigate the genes that are differently expressed in response to spaceflight in each genotype and to explore the similarities and differences of DEGs across the genotypes. However, a limited sample size (n = 3) alongside random distribution of experimental variation (error) can lead to random masking of the affected genes in each genotype group. In addition, lower power would be expected in genotypes with higher intra-group variation. These limitations can impact the identification of common DEGs affected by spaceflight across genotypes. Here, we provide a complementary analysis of the GLDS-120 dataset to evaluate the response of Arabidopsis root-tip to spaceflight across the dataset in light or dark conditions. To achieve this aim, we applied paired analyses of genotypes, which, through controlling for genotypes, allows the pooling of the samples to increase the sample size (n = 9), then used the resulting DEGs for gene ontology to identify the affected biological processes. In addition to providing a complementary analysis approach, to our knowledge, this is the first study to use gene set enrichment analysis to assess the impacted biological processes in Arabidopsis root-tip in spaceflight.

Results

The Arabidopsis samples within spaceflight and ground control groups were pooled and compared for differential gene expression. The outline of the analysis is shown in Supplemental Fig. 1. The original study defined the root-tip as the last 2 mm of root in the light samples and the last 1mm in the dark group (Paul et al. 2017). This difference in definition is likely due to the distinct development of roots in light and dark settings. Due to the difference in the definition of root-tip, we did not directly analyze the differential expression of genes between light and dark conditions.

Fig. 1
figure 1

Differential expression analysis of response to spaceflight in light conditions. A The distribution of samples across genotypes and locations using the top 500 most variable genes after regularized log transformation. B Volcano plot of the differential expressed genes between spaceflight and ground control with ground control as the reference group. Red points represent the 313 up-regulated genes; green points represent the 309 downregulated genes. Genes with adjusted p values < 0.05 are shown in color. Labels for the top 25 differentially expressed genes are provided. C Heatmap of differentially expressed genes between spaceflight and ground control groups. W Wassilewskija, C Columbia-0, P Columbia-0 PhyD, GC ground control, SF spaceflight, PC principal component

Full size image

Differentially expressed genes in light conditions during spaceflight

The summary of differential expression patterns of genes in the light is shown in Fig. 1. Principal component analysis (PCA) of the top 500 most variable genes showed a separation of genotype W from P and C along the first principal component axis (Fig. 1A), which is expected considering the genetic proximity of P and C genotypes. Interestingly, there also seem to be indications of separation of spaceflight and ground control conditions along the second principal component axis. Six hundred twenty-two differentially expressed genes (DEGs) were identified by comparing spaceflight and ground control groups using an adjusted p value cutoff of 0.05, with 313 upregulated and 309 downregulated genes in spaceflight. Expression of these DEGs, their respective p values, and their relative expression across samples can be seen as a volcano plot and a heatmap (Fig. 1B, C). The top 25 DEGs in light conditions based on their adjusted p values are provided in Supplemental Table 1.

Biological processes affected in light conditions during spaceflight

Gene ontology (GO) enrichment analysis using identified DEGs and an adjusted p value cutoff of 0.05 identified 37 enriched biological processes plotted in Fig. 2, with a prominent presence of terms related to the circadian cycle, light and photosynthesis, and hypoxia. The heatmap of DEGs in each cluster can be seen in Supplemental Figs. 2 and 3. To provide a simplified and more tangible overview of key affected genes and their interactions within terms, we also mapped our DEGs onto a selection of KEGG pathways. The “plant hormone signal transduction” pathway is included in Supplemental Fig. 4. We also included the KEGG plots for “circadian cycle”, “photosynthesis”, and “flavonoid biosynthesis” pathways due to their vicinity to enriched GO terms (Fig. 3; Supplemental Figs. 5, 6, and 7). Additionally, we checked the presence of our DEGs in a selection of KEGG pathways related to carbon metabolism (Supplemental Document, Supplemental Figs. 8 and 9).

Fig. 2
figure 2

Treeplot of the 37 identified biological processes using gene ontology enrichment analysis of the differentially expressed genes during spaceflight in light conditions. The terms were clustered using ‘average’ clustering. Seven groups were selected to improve the visualization of overlapping genes in Supplemental Figs. 2 and 3. The ratio is the percentage of differentially expressed genes compared to the genes in each term that were included in the differential expression analysis

Full size image
Fig. 3
figure 3

Mapping DEGs from light conditions onto KEGG A “circadian rhythm” B “photosynthesis” pathways. For the gene-wise breakdown of the differential expression results used for plotting the pathways, see Supplemental Figs. 5 and 6

Full size image
Fig. 4
figure 4

Differential expression analysis of response to spaceflight in dark conditions. A The distribution of samples across genotypes and locations using the top 500 most variable genes after regularized log transformation. B Volcano plot of the differential expressed genes between spaceflight and ground control with ground control as the reference group. Red points represent the 165 up-regulated genes; green points represent the 35 downregulated genes. Genes with adjusted p values < 0.05 are shown in color. The lower plot shows the selected region within the dotted rectangle from the top plot. Labels for the top 25 differentially expressed genes are provided. C Heatmap of differentially expressed genes between spaceflight and ground control groups. W Wassilewskija, C Columbia-0, P Columbia-0 PhyD, GC ground control, SF spaceflight, PC principal component

Full size image
Fig. 5
figure 5

Treeplot of the 13 enriched biological processes using gene ontology enrichment analysis of the differentially expressed genes during spaceflight in dark conditions. The terms were clustered using ‘average’ clustering. Five groups were selected to improve the visualization of overlapping genes in Supplemental Fig. 10. The ratio is the percentage of differentially expressed genes compared to the genes in each term that were included in the differential expression analysis

Full size image

Differentially expressed genes in dark conditions during spaceflight

The summary of differential expression patterns of genes in the dark is shown in Fig. 4. Similar to light conditions, the principal component analysis of the top 500 most variable genes showed a separation of genotype W from P and C along the first principal component axis (Fig. 4A). However, unlike light conditions, we did not notice a general separation of ground control and spaceflight samples within the first two principal components. A total of 200 DEGs were identified by comparing spaceflight and ground control groups using an adjusted p value cutoff of 0.05, with 165 upregulated and 35 downregulated genes in spaceflight. Expression of these DEGs, their respective p values, and their relative expression across samples can be seen as a volcano plot and a heatmap (Fig. 4B, C). The top 25 DEGs in dark conditions based on their adjusted p values are provided in Supplemental Table 2.

Biological processes affected in dark conditions during spaceflight

GO enrichment analysis using identified DEGs and an adjusted p value cutoff of 0.05 identified 13 enriched biological processes. These terms are plotted in Fig. 5, with a prominent presence of terms related to hypoxia among them. The heatmap of DEGs in each cluster can be seen in Supplemental Fig. 10. In addition, DEGs were mapped onto the “plant hormone signal transduction” KEGG pathway (Supplemental Fig. 11).

Discussion

Top 25 differentially expressed genes during spaceflight in light and dark conditions

A more detailed discussion of the top 25 differentially expressed genes in light and dark conditions can be found in the Supplemental Document. Briefly, thirteen of the top 25 DEGs in light conditions are upregulated during spaceflight, which include the genes with reported involvement in stress conditions, including abscisic acid response (ABF1 and NIC3), salicylic acid response (NIMIN-1), cesium (CML47 and AT3G28510) and zinc (AT1G29100) toxicity and potassium starvation (AT2G30660). Among the downregulated genes in light are genes involved in the circadian system and light signaling pathways (CCA1, LHY, RVE1, RVE2, RVE8, AT3G54500, CDF2, and HYH) and two genes involved in flavonoid biosynthesis (CHS and CHI3). These observations suggest that stress responses, the circadian system, and flavonoid synthesis are affected during spaceflight in the light.

Similar to light conditions, several of the 25 most significant genes in dark conditions have been previously associated with various stress responses, including among the upregulated genes (RCI2A, XERO1, CXXS1, WRK70, HIS1-3, MBF1C, WRKY45, AT2G05510, NIMIN-1 and RAP2-9) and downregulated genes (MS2, AAP4 and TOPP7). SWEET13 and SWEET14 are also among the top downregulated genes in dark conditions. The involvement of SWEET13 and SWEET14 in cellular uptake of gibberellin and root development has been previously reported (Kanno et al. 2016).

Enriched biological processes during spaceflight

There is a prominent enrichment for hypoxia-related terms in spaceflight, with 6 of the 37 enriched terms in the light and 6 of the 13 enriched terms in the dark belonging to these terms (Figs. 2, 5). These findings suggest that hypoxia remains a key challenge for root development in spaceflight conditions, and engineering solutions on this front can improve plant growth in spaceflight. Identifying hypoxia-related terms is consistent with the well-documented hypoxic metabolic response of the root system in spaceflight (Porterfield 2002) and is likely due to convection inhibition in microgravity (Liao et al. 2004). The term "cellular response to hypoxia" was also enriched in studying the response to microgravity in whole seedlings (Villacampa et al. 2021).

Twenty-one of the 37 enriched terms in light conditions are related to light signaling, photosynthesis, or circadian processes. Consistently, enrichment for processes related to photosynthesis and response to light has been previously observed in whole root samples (Califar et al. 2020). The distribution of DEGs that drive the enrichment for light biological terms is plotted in Supplemental Fig. 3. Positive phototropism of root to blue light and red light has been observed in the absence of gravity and is suppressed in gravity (Vandenbrink et al. 2016). Photomorphogenesis of the root is regulated through interaction with the shoot and through the root’s own photoreceptors (Van Gelderen et al. 2018). It has been demonstrated that the direct exposure of the root to light can enhance its gravitropism (Burbach et al. 2012; Suzuki et al. 2016). Our current results suggest that the interaction between light signaling and gravitropism is likely reciprocal, with the absence of gravity also affecting the root’s phototropism. While our analysis cannot systematically determine the direction of change in the light signaling pathway, the expression of genes encoding transcription factors HY5 and HYH are both significantly downregulated in our results (Supplemental Fig. 5). HY5 is a key regulator of photomorphogenesis (Gangappa and Botto 2016) and, alongside HYH, modulates the development and gravitropism of lateral root (Sibout et al. 2006), which suggests microgravity might have a detrimental impact on the root’s photomorphogenesis. Enrichment for the circadian terms is not surprising considering their close interaction with components of the light signaling terms (Fig. 3A) (Más 2008). Interestingly, the circadian process has also been affected in Arabidopsis rosette leaves during spaceflight (Wang et al. 2022). However, in contrast to the current study, the key genes RVE1, RVE2, LHY, and CCA1 were reported to be upregulated in these experiments. There are key differences in the experiment design between the two studies, including different spaceflight environments and differences in the age of studied plants. Therefore, whether the opposing direction of gene expression results from tissue-specific responses to spaceflight remains to be determined.

Enrichment for photosynthesis-related terms is of interest as it suggests their components might be active in the root-tip in the reported conditions. Importantly, Vandenbrink et al. observed the downregulation of genes involved in several light signaling-related pathways, including “photosynthesis” and “porphyrin and chlorophyll metabolism” through analysis of whole seedlings’ transcriptomic response to spaceflight (Vandenbrink et al. 2019). Ten of the 11 gene sets with at least one downregulated member in our KEGG photosynthesis pathway (Fig. 3B) also show downregulation in the results by Vandenbrink et al., indicating a robust consistency across the two studies. This observation also suggests that the downregulation of photosynthesis components might not be tissue-specific and is potentially a systemic response to spaceflight. The downregulation of genes involved in the photosynthetic pathway is in line with an array of studies describing the negative impact of spaceflight on photosynthesis in other organisms (Vandenbrink et al. 2019). Whether the differential expression of photosynthesis components indicates active photosynthesis in the root under the reported conditions is unknown. Since activation of photosynthesis pathways in the root is not expected under field conditions, the contributing genes might be ectopically expressed due to the exposure of the root to light. Should the involvement of photosynthesis processes in the root be evaluated as significantly detrimental to the development of plants in spaceflight conditions, design improvements in line with the dark agar plates and D-root design could be considered in future missions (Xu et al. 2013; Silva-Navas et al. 2015).

Significant enrichments in our results for "response to auxin", "flavonoid metabolic process", and "flavonoid biosynthetic process" are in line with the role of auxin and flavonoids in root’s gravitropism (Buer and Muday 2004; Su et al. 2017). Enrichment for auxin signaling in the analysis of seedlings in spaceflight (Paul et al. 2012) and elevation in isoflavone glycosides in response to microgravity in soybean root (Levine et al. 2001) have been reported. The observed enrichment for "response to hydrogen peroxide" is in line with the role of hydrogen peroxide in root differentiation (Dunand et al. 2007). Elevated hydrogen peroxide in response to microgravity in a cell culture model has also been reported (Hausmann et al. 2014). Enrichment for "response to salicylic acid" is consistent with the salicylic acid pathway's involvement in abiotic responses (Khan et al. 2015). Additionally, enrichment for "regulation of abscisic acid-activated signaling pathways" might be related to the reports of "response to abscisic acid" among the enriched terms from whole seedlings analysis during spaceflight (Villacampa et al. 2021).

In dark conditions, enrichment for "systemic acquired resistance" is interesting, considering the pathway's role in response to abiotic stress and plant development (Klessig et al. 2018). Whether the enrichment for "response to ethylene" in the dark is related to the potential accumulation of ethylene in the International Space Station remains to be determined. Elevated levels of atmospheric ethylene were implicated as a cause of developmental differences observed in spaceflight during the shuttle spaceflight program (Guisinger and Kiss 1999; Kiss et al. 1999).

An intrinsic limitation of gene ontology analysis is the enrichment of terms that might seem less relevant to the tissue of study. This could be due to the involvement of the term genes in other functions or ectopic expression of genes. For instance, in our analysis of dark conditions, enrichment for “seed growth” and “anther dehiscence” seems to be driven by differential expression of the SWEET13 and SWEET14 genes, which, as pointed out earlier, are also expressed in the root and are involved in the cellular uptake of gibberellin, potentially impacting cell division, root elongation, and root-tip environmental response kinetics (Kanno et al. 2016).

Due to the potential significance of carbon metabolism in spaceflight response in plants undergoing photosynthesis, we checked the presence of DEGs in a selection of pathways involved (Supplemental Document). Several genes listed in the KEGG "Glycolysis/Gluconeogenesis" and "Starch and sucrose metabolism" pathways are among our DEGs (Supplemental Figs. 8–9). It is, however, important to note that the terms directly related to these pathways were absent in our gene set enrichment results.

Comparison of current findings with the parent study

The original study by Paul et al. (2017) evaluated differentially expressed genes (DEGs) within each genotype and investigated the overlapping DEGs across genotypes. In our analysis, we use the dataset to identify the transcriptional response of Arabidopsis root-tip to spaceflight independent of the genetic differences. Therefore, the two studies use different analysis approaches to answer different underlying questions.

To evaluate the general consistency of our findings against the parent study, we compared the list of our DEGs with the common transcripts identified by Paul et al. (2017). Fourteen of the 18 common DEGs identified by Paul et al. in light conditions have unique TAIR IDs and were also among our DEGs, with 10 of them within the top 25 DEGs (Supplemental Table 3). Another gene (AT2G43920, HOL2) was excluded in our pipeline due to DESeq’s filtration of genes that contain extreme outliers from the analysis, but the gene appears to show a strong change in expression (2.7 log2 fold change) during the spaceflight. The remaining three of the 18 genes were each aligned to two reference TAIR IDs, with at least one of the two reference genes listed as a target in our analysis. In dark conditions, 6 of the 11 common genes identified by Paul et al. were also among the DEGs in our results, with four among the top 25 DEGs. We believe the partial overlap of targets in dark conditions is likely due to the apparent weaker response to spaceflight in the dark setting. This could allow for the differences in analysis pipelines to have more pronounced impacts on the results.

Paul et al. presented their list of common DEGs in functional groups, with 11 of the 18 common DEGs in the light and 10 of the 11 common DEGs in the dark grouped as “associated with defense or wall”. These are broadly consistent with our enriched terms, which show various stress responses in light conditions and enrichment for “systemic acquired resistance” in dark conditions. The remaining seven common DEGs in the light were grouped as “associated with light signaling”, consistent with our strong enrichment for light and circadian terms. While we report different analysis goals and approaches in the current study, our findings are broadly in line with the parent study.

Study limitations

Engineering components determine many of the key environmental variables during spaceflight. Considering that most available transcriptional studies have been conducted on board the International Space Station, whether the identified responses are truly universal or vessel specific remains to be determined. Comparison of current findings with those of the upcoming Moon to Mars missions or experiments by other international players could significantly improve the identification of truly universal responses to spaceflight.

We also noticed a pronounced variation across samples within each experimental group, as shown in our PCA plots. Automated pipelines in both the experimental stage and sample preparation steps could significantly reduce any contributing technical variations, increase the reproducibility within experiment groups, and improve the quality of transcriptomics data and analysis power.

Summary

Our complementary reanalysis of the GLDS-120 dataset identified robust transcriptional responses and arrays of enriched biological terms as part of Arabidopsis root-tip response to spaceflight in light and dark conditions. We identified 622 DEGs in light conditions and 200 DEGs in dark conditions. Terms related to hypoxia response were enriched in both conditions during spaceflight. Furthermore, there was a pronounced enrichment for light and circadian terms in light conditions and various stress responses in both dark and light conditions. Our study also demonstrates the scientific merits of open-source data sharing through databases like the NASA GeneLab. Space biology data is sparse and difficult to produce firsthand because conducting biology experiments in spaceflight requires specialized equipment and is even more expensive than conducting terrestrial experiments. We note, therefore, that data sharing is beneficial for optimizing scientific advancement in a field like space biology. Our findings further the current knowledge regarding the molecular response of plants to spaceflight conditions and, in turn, can pave the way for improvements in environmental engineering and the choice of plants in biological life support systems.

Methods

The methods section is provided as a supplemental document.