PhoH2 proteins are found in a diverse range of organisms that span the bacterial tree and little is known about this large protein family. PhoH2 proteins have two domains: An N-terminal PIN domain fused to a C-terminal PhoH domain. The genome of Mycobacterium tuberculosis encodes 48 PIN domains and 47 of these constitute the VapC components of the 47 VapBC toxin-antitoxins. The 48th member of the M. tuberculosis PIN domain array is found in the single PhoH2 protein encoded in the genome. All characterized PIN domain proteins are RNases and the PhoH domains are predicted ATPases. This fusion of a PIN domain with an ATPase reflects a much wider association between PIN domains and PhoH domains across many prokaryote genomes. Here, we examine PhoH2 proteins from M. tuberculosis, Mycobacterium smegmatis and a thermophilic homologue from Thermobispora bispora and we show that PhoH2 is a sequence-specific RNA helicase and RNAse. In addition, phoH2 from M. tuberculosis and M. smegmatis is part of a longer mRNA transcript which includes a small, unannotated open reading frame (ORF) upstream of the phoH2 gene. This small gene overlaps with the beginning of the phoH2 gene in a manner similar to the PIN domain toxin-antitoxin operons. We have annotated the upstream gene as phoAT and its putative promoter elements satisfy previously characterized consensus sequences at the −10 site. Conditional growth experiments carried out in M. smegmatis revealed a negative effect on growth by the expression of M. tuberculosis PhoH2 that was alleviated by co-expression of the PhoAT peptide. Thus in M. tuberculosis, PhoH2 represents a new variation on a type II PIN domain toxin-antitoxin systems such that the toxin-antitoxin is now coupled to an RNA helicase whose predicted biological function is to unwind and cleave RNA in a sequence specific manner.