Polymerase delta-interacting protein 38 (PDIP38) modulates the stability and activity of the mitochondrial AAA+ protease CLPXP

Abstract

Over a decade ago Polymerase δ interacting protein of 38 kDa (PDIP38) was proposed to play a role in DNA repair. Since this time, both the physiological function and subcellular location of PDIP38 has remained ambiguous and our present understanding of PDIP38 function has been hampered by a lack of detailed biochemical and structural studies. Here we show, that human PDIP38 is directed to the mitochondrion in a membrane potential dependent manner, where it resides in the matrix compartment, together with its partner protein CLPX. Our structural analysis revealed that PDIP38 is composed of two conserved domains separated by an α/β linker region. The N-terminal (YccV-like) domain of PDIP38 forms an SH3-like β-barrel, which interacts specifically with CLPX, via the adaptor docking loop within the N-terminal Zinc binding domain of CLPX. In contrast, the C-terminal (DUF525) domain forms an immunoglobin-like β-sandwich fold, which contains a highly conserved putative substrate binding pocket. Importantly, PDIP38 modulates the substrate specificity of CLPX and protects CLPX from LONM-mediated degradation, which stabilises the cellular levels of CLPX. Collectively, our findings shed new light on the mechanism and function of mitochondrial PDIP38, demonstrating that PDIP38 is a bona fide adaptor protein for the mitochondrial protease, CLPXP.

Crystal structure and molecular dynamics of human POLDIP2, a multifaceted adaptor protein in metabolism and genome stability

“…β0 hydrogen bonds with the distal β15 strand of the upper β‐sheet (Figure 3(b)), thereby connecting DUF525 to the N‐terminal region (Figure 2(a)). This is in contrast to Strack et al, 16 where a few N‐terminal ordered residues ahead of the β1 strand point away from the main body of the structure and β0 is not observed.…”

“…The POLDIP2 51‐368 structure was solved with molecular replacement, using bacterial HspQ (YccV) 14 and human FBxo3 (DUF525), 15 with one molecule in the asymmetric unit and R work / R free factors of 0.216/0.288 respectively, 99.67% completeness and 99.6% of residues in preferred or allowed Ramachandran regions (Table 1, Figure S3(c)). During preparation of this manuscript, another structure of POLDIP2 was independently reported, albeit of a lower resolution (3.4 Å) 16 . This was generally consistent with our structure, although here we also report number of additional observations.…”

“…During preparation of this manuscript, another structure of POLDIP2 was independently reported, albeit of a lower resolution (3.4 Å). 16 This was generally consistent with our structure, although here we also report number of additional observations. The POLDIP2 crystal structure exhibits a globular fold, with the YccV-like and DUF525 domains juxtaposed on top of each other (Figure 2(a)).…”

“…Although another POLDIP2 structure was reported during preparation of this manuscript, 16 the data presented here offer additional insights and features not observed in the previous analysis. This includes the presence of a channel traversing the POLDIP2 core which we speculate could house a conformational switch, owing to a putative modification of Cys 266 contained within the channel and its location at the E‐YccV/DUF525 domain interface.…”

“…During structural refinement we observed that Cys 266 exhibited electron density that could not be accounted for, either by alternative rotamers or by surrounding residues (green, Figure S4, left panel). This additional density was not observed in Strack et al, 16 presumably following the lower resolution dataset, although a direct comparison cannot be made as the structure had not been released at the time of writing (PDB : http://firstglance.jmol.org/fg.htm?mol=6ZLX). We propose this represents an in crystallo cysteine modification, as DNA sequencing confirmed that the electron density at this position did not result from a residue substitution mutation.…”