In this section we review the GO terms that were specifically created by the PAMGO project for secretion systems. Many of the functions and processes of proteins related to secretion systems (for example effectors) can be described with GO terms from other parts of the GO hierarchy; those are not covered here in detail. We also note that many additional terms are still needed in this area, especially for secretion systems that are not central to bacteria-host interactions and which therefore have received less attention from the PAMGO consortium.
The following schema illustrates the parallel nature of the GO terms for each of the secretion systems:
• type (I–IV, VI) protein secretion system complex (Type VII currently missing; type V does not exist as a distinct identifiable complex)
• children of "GO:0009306 protein secretion":
o protein secretion by the type (I–VI) secretion system (Type VII missing)
• descendants of "GO: 0052047 interaction with other organism via secreted substance during symbiotic interaction"
o "interaction with host via protein secreted by type II/III/IV secretion system"
o "modification of morphology or physiology of other organism via protein secreted by type II/III/IV secretion system during symbiotic interaction"
o both of these have as child:
"modification by symbiont of host morphology via protein secreted by type II/III/IV secretion system"
The terms above are depicted in Figure . Note that in the component ontology and among the children of "GO:0009306 protein secretion" there is just one term for each secretion system; hence the use of such terms is straightforward and perfectly parallel for all secretion systems that have been addressed so far by the PAMGO consortium. Currently, detailed descendant terms of "GO: 0052047 interaction with other organism via secreted substance during symbiotic interaction" are available only for systems II, III, and IV. However, as noted in the survey of secretion systems above, examples exist in which organism interactions are modulated by proteins secreted via systems I, V, VI and VII as well as via the universal Sec and Tat pathways. Thus the PAMGO consortium is currently creating parallel terms for these six systems. Note also that no system-specific terms have yet been created in the molecular function ontology.
Figure 2 Gene Ontology terms for secretion systems under "cellular component" and "biological process.". GO terms for secretion systems described in this review article are encased in dashed boxes. (A) shows terms that are children of the process term "GO ID: (more ...)
The family of terms "Interaction with host via protein secreted by type number secretion system" is appropriate for annotating gene products that form the apparatus of secretion when there is experimental evidence that the interaction with the host is affected by secretion through that apparatus. As an example (once terms for the T7SS have been created), in mycobacterial pathogens that contain multiple T7SS gene clusters, if deletion of a cluster affected virulence then the gene in the cluster could be annotated with "Interaction with host via protein secreted by type VII secretion system". However, if deletion of a different cluster did not affect virulence then the term would not be appropriate for that cluster and only the term "protein secretion by the type VII secretion system" would be appropriate. Similarly, in a non-pathogen such as Mycobacterium smegmatis, none of the T7SS systems should be annotated with "Interaction with host via protein secreted by type VII secretion system", even if a particular T7SS gene cluster were orthologous to a T7SS in a pathogenic species. On a similar theme, if experimental evidence shows that a gene or gene cluster is important to symbiosis, it may be annotated with "Interaction with host via protein secreted by type number secretion system", even if some genes in the cluster appear to be pseudogenes; thus experimental evidence takes precedence over bioinformatic inferences.
The family of terms "modification of morphology or physiology of other organism via protein secreted by type number
secretion system during symbiotic interaction" and "modification by symbiont of host morphology via protein secreted by type number
secretion system" are appropriate for annotating the effector proteins that are transported by the secretion systems, but not for the components of the secretion system itself. On the other hand, there are many cases where proteins have a dual function as part of the transport machinery and as effectors. The most striking of these is the "autotransporter" proteins that are secreted via the T5SS pathway in which an N-terminal effector domain is fused to a C-terminal transporter domain. Some proteins associated with the T6SS also appear to be similarly bi-functional [38
A common theme among most of the secretion systems is the role of ATP hydrolysis and chaperones (Figure ). This is not yet captured in a systematic way in the GO. Nevertheless the following terms are appropriate in this context: "GO: 0015450 P-P-bond-hydrolysis-driven protein transmembrane transporter activity" and "GO: 0016887 ATPase" and "GO:0042623 ATPase activity, coupled", while "GO: 0043190 ATP-binding cassette (ABC) transporter complex" would be appropriate for the T1SS.
The T2SS and T5SS (and in certain cases T4SS and T1SS as well) deserve a special note because of their relationship with the Sec and Tat pathways. As noted in the first part of this article, proteins translocated via T2SS or T5SS (and sometimes the T1SS and T4SS) first go through the Sec or the Tat pathways. GO provides two pairs of parallel terms for the component and process aspects of the Sec and Tat pathways. "GO:0031522 cell envelope Sec protein transport complex" (component) and "GO:0043934 protein transport by the Sec complex" (process) are available for the Sec pathway; and "GO:0033281 Tat protein transport complex" (component) and "GO:0043935 protein transport by the Tat complex" (process) are the corresponding terms for the Tat pathway. As an example, a protein that affects the interaction with a host that is eventually translocated by the T2SS but goes through the Sec pathways first, could be annotated using both "GO: 0052051 Interaction with host via protein secreted by type II secretion system" and "Interaction with host via protein secreted by Sec secretion system"