Short, Rich, and Powerful: a New Family of Arginine-Rich Small Proteins Have Outsized Impact in Agrobacterium tumefaciens

Due to minute size and limited sequence complexity, small proteins can be challenging to identify but are emerging as important regulators of diverse processes in bacteria. In this issue of the Journal of Bacteriology, Kraus and coworkers (A. Kraus, M. Weskamp, J. Zierles, M. Balzer, et al., J Bacteriol 202:e00309-20, 2020, https://doi.org/10.1128/JB.00309-20) report a comprehensive analysis of a fascinating subfamily of arginine-rich small proteins in Agrobacterium tumefaciens, conserved among Alphaproteobacteria.

ABSTRACT

Due to minute size and limited sequence complexity, small proteins can be challenging to identify but are emerging as important regulators of diverse processes in bacteria. In this issue of the Journal of Bacteriology, Kraus and coworkers (A. Kraus, M. Weskamp, J. Zierles, M. Balzer, et al., J Bacteriol 202:e00309-20, 2020, https://doi.org/10.1128/JB.00309-20) report a comprehensive analysis of a fascinating subfamily of arginine-rich small proteins in Agrobacterium tumefaciens, conserved among Alphaproteobacteria. Their findings reveal that these small proteins are under complex regulation and have a disproportionately large impact on metabolism and behavior.

TEXT

The issue of scale has a major impact on our ability to recognize important constituents in the world around us. Objects of small scale may have a tremendous impact but can be easily obscured by the larger context in which they reside. An example of this phenomenon, among many, is the relatively meager recognition and understanding of small proteins compared to that of their larger, more readily identifiable counterparts. Small proteins are encoded by short coding sequences that specify 50 amino acids or fewer and often lack clearly annotated domains or motifs indicative of their structure or function. This makes identification and characterization of genes for small proteins in genomic sequences more challenging than for larger proteins. Recent advances have led to the prediction and initial characterization of a growing number of small proteins in Escherichia coli and other bacteria (1, 2). Emerging evidence suggests that, often, small proteins interact with cellular membranes or protein complexes and can play influential roles in regulating diverse processes, such as sugar transport, protection from phage invasion, and regulation of sensor kinase function (2–4).

A related but even broader challenge is the prediction of protein function from sequence information. Even for larger proteins possessing predicted secondary structure motifs and recognized domains, a challenge remains in characterizing genes with domains of unknown function (DUFs). More than 20% of protein families in the Pfam database are annotated as DUFs (5). Many DUFs are highly conserved and may serve important or even essential functions (6), but more studies into the structure and function of these domains are needed to understand their cellular roles.

Recent work has identified a family of conserved arginine-rich proteins, classified as DUF1127, that are typically quite short (7). Almost all DUF1127 domain-containing proteins are composed of single domains and are fewer than 100 amino acids (aa) in length, with approximately 15% of these under 51 aa, that can be classified as true small proteins. DUF1127 is found primarily in members of the Alphaproteobacteria and Gammaproteobacteria (7, 8). The Gammaproteobacteria typically harbor a single DUF1127 protein, annotated as YjiS, the function of which has not been reported (9).

In this issue of the Journal of Bacteriology, Kraus et al. report compelling initial analysis of multiple DUF1127-containing proteins in the plant pathogen Agrobacterium tumefaciens, providing new insights into their functions and distribution among the members of the Alphaproteobacteria (8). There have been a few prior studies on the Alphaproteobacteria DUF1127 proteins. The DUF1127 protein RSP_6037 in Rhodobacter sphaeroides is in an operon with and affects expression of four small RNAs (sRNAs), with the whimsical name of the cuckoo RNAs (CcsR1 to CcsR4 [CcsR1-R4]) for their conserved sequence (CCUCCUCCC) motif (10, 11). The CcsR RNAs and RSP_6037 regulate C₁ carbon metabolism (10). Expression of RSP_6037 and another DUF1127-containing gene, RSP_0557, are induced in the transition from exponential to stationary phase in the presence of oxygen (12). In Brucella abortus, three DUF1127 proteins are activated by VtlR (13), a LysR-type regulatory protein that is required for virulence (14).

Rather than relying on genome annotations, which often overlook small proteins (1), Kraus et al. (8) took an unbiased approach to identify Agrobacterium tumefaciens proteins containing the DUF1127 domain. A. tumefaciens is a facultative plant pathogen and the causative agent of crown gall disease (15, 16). To find potential A. tumefaciens DUF1127 proteins, Kraus et al. (8) translated all possible reading frames from the A. tumefaciens C58 genome sequence and used BLAST to search the in silico amino acid sequence against known DUF1127 proteins (17). This targeted approach led to the identification of seven proteins in A. tumefaciens containing the arginine-rich sequence of DUF1127. Of these seven, four were previously annotated to contain DUF1127, two were listed as hypothetical proteins of unknown function, and one was encoded by a newly annotated open reading frame (ORF) that had formerly been described as an sRNA. Three of the A. tumefaciens DUF1127 proteins are true small proteins (<50 aa), and are designated short DUF1127 proteins (SDPs), and the other four are considered long DUF1127 proteins (LDPs), ranging from 72 to 101 aa. Kraus et al. (8) observed that the short and long DUF1127 proteins in A. tumefaciens appeared to fall into distinct protein groups and asked whether other DUF1127 proteins in the InterPro database fall into similar subdivisions (7). Indeed, they found three distinct subclasses of DUF1127 proteins: those similar to the A. tumefaciens SDPs that were found mostly in Alphaproteobacteria, those similar to the E. coli YjiS protein and largely restricted to the Gammaproteobacteria, and a third group, including the A. tumefaciens LDPs distinct from the first two categories. Kraus et al. (8) also observed that in the Alphaproteobacteria, the DUF1127 SDP proteins are almost always associated with a gene encoding an LsrB-like regulatory protein and cuckoo sRNAs. LsrB is a LysR-type protein and is homologous to VltR from B. abortus. Prior studies on the A. tumefaciens LsrB protein found that it stimulates biofilm formation and attachment to host plants to drive infection, and the ΔlsrB mutant is deficient in plant transformation (18). Kraus et al. (8) tested whether deletion of lsrB affects DUF1127 protein expression and observed that the absence of LsrB decreases expression of the SDPs and reciprocally increases LDP expression. During growth in culture, each SDP is preferentially expressed at different stages of the growth curve, but they all appear to be induced at elevated temperature, suggesting some common regulation as well as some specific control of each SDP.

Kraus et al. (8) investigated the role for the SDPs in A. tumefaciens physiology. Although individual deletion of the SDP genes had little obvious phenotypic effect, simultaneous deletion of all three SDP genes resulted in a profound growth defect during the transition to stationary phase. The differential in growth between the triple SDP mutant and the wild type was dependent on levels of both sucrose and NaCl in the medium, reflecting potential connections to metabolism and perhaps osmotic stress. Kraus et al. (8) looked for changes in protein abundance between wild-type and the triple SDP mutant cell extracts by one-dimensional SDS-PAGE analysis. They observed an increase in one obvious band in protein extracts from the triple SDP mutant in late exponential phase, which they identified through mass spectrometry to contain the phosphate ABC transporter component PtsS and NADP-dependent alcohol dehydrogenase Adh. Strong induction of the corresponding genes in the triple SDP mutant was confirmed by Northern blot analysis. Transcriptome sequencing (RNA-seq) analysis comparing the transcripts of the wild-type strain to that of the triple SDP mutant revealed that, strikingly, thousands of genes were differentially expressed in the triple SDP mutant at various time points in the growth cycle. Among the genes upregulated in the triple SDP mutant at all growth phases were the soxBDAG and glyA genes involved in glycine/serine homeostasis and C₁ metabolism and the norDBC genes involved in denitrification. Regulation of carbon metabolism has been linked to SDPs in other Alphaproteobacteria, as deletion of a B. abortus SDP gene affects fucose utilization, and in R. sphaeroides, overexpression of the cuckoo sRNAs with an SDP modulates C₁ metabolism (10, 13). Genes associated with the phosphorus starvation response, including ptsS and phoB, were upregulated in the triple SDP mutant in late exponential phase. Two LDP genes were upregulated in the triple SDP mutant, suggesting a genetic interaction between the two classes of DUF1127 proteins.

Deletion of all three SDPs led to an increase in cell-cell aggregation and biofilm formation. Colonies of this mutant are elevated for exopolysaccharides that bind the dye Congo red (predominantly cellulose and the unipolar polysaccharide) and decreased for binding of calcofluor white, indicative of decreases in succinoglycan (19, 20). This pattern of polysaccharide production is also consistent with elevated biofilm formation in A. tumefaciens. The RNA-seq results suggested that the triple SDP mutant mimics cells undergoing phosphorus limitation, a condition that has been shown to trigger elevated attachment and biofilm formation in A. tumefaciens, and may be at the root of this observed phenotype (21, 22). A. tumefaciens is well known as a plant pathogen that genetically transforms plants by DNA transfer (16). Kraus et al. (8) report that when assayed using a qualitative Arabidopsis seedling infection test, the triple SDP mutant retained A. tumefaciens virulence; in contrast, the transcriptional activator of the SDPs, LsrB, is required for efficient transformation (18). Therefore, LsrB likely regulates additional cellular components that influence plant transformation and virulence. This result is reminiscent of data from B. abortus suggesting that, while the regulator protein VtlR is required for virulence, deletion of the SDPs does not significantly influence B. abortus virulence (13). There may be yet-undiscovered conserved LsrB (VtlR)-activated genes or pathways that participate more directly in these virulence programs. Deletion of all the three SDP genes led to increased expression of the cuckoo sRNA L5, immediately downstream of the gene for SDP3, suggesting a genetic interaction between the small proteins and the sRNA. The roles of this and the other cuckoo sRNAs in A. tumefaciens remain unclear.

Kraus et al. (8) provide new insight into the different classes of DUF1127 proteins in bacteria and, particularly, the role of the SDP-type DUF1127 proteins in A. tumefaciens. This study lays the groundwork for investigations into the mechanism by which these tiny proteins impact large-scale phenotypes such as growth, metabolism, and biofilm formation. While the single SDP mutants did not have strong phenotypes, suggesting redundancy, the loss of all three SDP proteins led to dramatic changes in growth, metabolism, and biofilm formation. It is interesting that the four LDPs of A. tumefaciens are oppositely impacted in the lsrB mutant relative to the SDPs, but the function of these proteins has not been characterized. The strong positive charge imparted through the arginine-rich sequence of DUF1127 proteins, particularly in the SDPs (8), may facilitate interactions with RNA, consistent with their proximity to and regulatory impact on the cuckoo sRNAs. Interestingly, the three SDPs of B. abortus localized to the membrane in cell fractionation experiments, and this may be a conserved attribute of the SDPs (13). The positive charge of the SDPs could also mediate interactions with negatively charged membrane phospholipids. These two potential interaction targets are not mutually exclusive. Future work should build from the current studies and reveal how these fascinating small proteins are integrated into the control of cellular physiology and how they impart their far-reaching effects in A. tumefaciens and other members of the Alphaproteobacteria.

The views expressed in this article do not necessarily reflect the views of the journal or of ASM.