Type IV pilus assembly proficiency and dynamics influence pilin subunit phospho-form macro- and microheterogeneity in Neisseria gonorrhoeae

Abstract

The PilE pilin subunit protein of the gonococcal Type IV pilus (Tfp) colonization factor undergoes multisite, covalent modification with the zwitterionic phospho-form modification phosphoethanolamine (PE). In a mutant lacking the pilin-like PilV protein however, PilE is modified with a mixture of PE and phosphocholine (PC). Moreover, intrastrain variation of PilE PC modification levels have been observed in backgrounds that constitutively express PptA (the protein phospho-form transferase A) required for both PE and PC modification. The molecular basis underlying phospho-form microheterogeneity in these instances remains poorly defined. Here, we examined the effects of mutations at numerous loci that disrupt or perturb Tfp assembly and observed that these mutants phenocopy the pilV mutant vis a vis phospho-form modification status. Thus, PC modification appears to be directly or indirectly responsive to the efficacy of pilin subunit interactions. Despite the complexity of contributing factors identified here, the data favor a model in which increased retention in the inner membrane may act as a key signal in altering phospho-form modification. These results also provide an alternative explanation for the variation in PilE PC levels observed previously and that has been assumed to be due to phase variation of pptA. Moreover, mass spectrometry revealed evidence for mono- and di-methylated forms of PE attached to PilE in mutants deficient in pilus assembly, directly implicating a methyltransferase-based pathway for PC synthesis in N. gonorrhoeae.

Figure 1. Disrupting pilus biogenesis results in PptA-dependent PC-modification of PilE.

A) Schematic representation of the PE and PC structures covalently bound by O-linkage to serine residues of pilin. B) Immunoblots of whole cell lysates made from equal numbers of cells and of equal amounts of protein from purified pili using the PC recognizing antibody TEPC-15 (upper panel) and the PilE peptide specific α-pilin antibody (lower panel). Strains used were wild-type (N400), pptA (KS9), pilE _S68A (KS640), pilD (KS641), pilD pptA (KS662), pilD pilE _S68A (KS667), pilF (KS643), pilF pptA (KS663), pilF pilE _S68A (KS668), pilQ (KS644), pilQ pptA (KS664), pilQ pilE _S68A (KS669), pilP (KS665), pilP pptA (KS666), pilP pilE _S68A (KS670), pilG (KS674), pilG pptA (KS673), and pilG pilE _S68A (KS672). All samples were run on the same gel and the dotted lines were introduced as guidance facilitating evaluation of the data. Results representative of at least three different experiments are shown.

Figure 2. Lack of pilus associated proteins leads to PptA-dependent PC modification of PilE.

Immunoblot analysis of whole cell lysates of equal numbers of cells and of equal amounts of protein from purified pili. The antibodies used were the PC recognizing TEPC-15 and the PilE peptide specific antibody α-pilin. Strains used were in A) wild-type (N400), pptA (KS9), pilT _ind (12/9/1), pilE _ind (MW24), pilC (KS787), pilC pptA (KS788), pilC pilT _ind (KS789), pilV (KS790), pilV pptA (KS10), pilV pilT _ind (KS791), comP (KS792), comP pptA (KS793), comP pilT _ind (KS794), pilU (KS795), pilU pptA (KS796), pilU pilT _ind (KS798) and in B) wild-type (N400), pilE _ind (KS786), pilH (KS799), pilH pptA (KS800), pilH pilT _ind (KS801), pilI (KS802), pilI pptA (KS803), pilI pilT _ind (KS804), pilJ (KS805), pilJ pptA (KS806), pilJ pilT _ind (KS807), pilK (KS808), pilK pptA (KS809), pilK pilT _ind (KS810), pilL (KS811), pilL pptA (KS812) and pilL pilT _ind (KS813). The faster migrating protein band below pilin is S-pilin (indicated by an arrow), a proteolytic degradation product of PilE that is a correlate of type IV pilus biogenesis defects and which requires pilT expression . The strains were grown on standard GC plates without inducer such that the pilT _ind and pilE _ind loci were not expressed. All samples on each blot were run on the same gel and the dotted lines were introduced as guidance facilitating evaluation of the data. Results representative of at least three different experiments are shown.

Figure 3. Mutations in pilE that perturb assembly of Tfp lead to PC modification of PilE.

Immunoblot of whole cell lysates of equal numbers of cells and of equal amounts of protein from purified pili. The antibodies used were the TEPC-15 antibody and the PilE peptide specific α-pilin antibody. – denotes a null allele and + denotes a wild-type allele of pptA. Strains used were wild-type (N400), pilE _ind (4/3/1), iga::pilE (KS130), iga::pilE pptA (KS813), iga::pilE _E5L (KS814), and iga::pilE _E5L pptA (KS815) iga::pilE _E5V (KS816), and iga::pilE _E5V pptA (KS817), iga::pilE _G1S (KS818), and iga::pilE _G1S pptA (KS819), iga::pilE _AAM38-40 (KS769), and iga::pilE _AAM38-40 pptA (KS821), iga::pilE _I4T (KS722), and iga::pilE _I4T pptA (KS723), iga::pilE _V9M (KS724), and iga::pilE _V9M pptA (KS774), iga::pilE _A20T (KS775), and iga::pilE _A20T pptA (KS776), iga::pilE _AAM38-40His (KS525), iga::pilE _AAM38-40His pptA (KS781), iga::pilE _E5KHis (KS784), and iga::pilE _E5KHis pptA (KS820). The strains were grown on standard GC plates without inducer such that pilE _ind was not expressed. All samples on each blot were run on the same gel and the dotted lines were introduced as guidance facilitating evaluation of the data. Results representative of at least three different experiments are shown.

Figure 4. Overexpression of PilE results in increased PC-modification.

Shown are immunoblot analyses of cell lysates made from equal numbers of cells of A) whole cells and B) cells recovered following shear depletion of pili. In whole cell lysates C) and D) the 2xpilE and 3xpilE strains were diluted 1∶1 and 1∶2 to account for the increased amount of PilE. The antibodies used were the PC recognizing antibody TEPC-15 and the PilE recognizing α-pilin antibody. – denotes a null allele and + denotes a wild-type allele of pptA. The strains used were the wild-type (N400) expressing one copy of pilE, 2xpilE (iga::pilE, i.e. a wild-type background expressing two copies of pilE) (KS646), 3xpilE (iga::2xpilE, i.e. a wild-type background expressing three copies of pilE) (KS647), and 3xpilE pptA (pptA iga::2xpilE, i.e. a pptA background expressing three copies of pilE) (KS653). Results representative of at least three different experiments are shown.

Figure 5. Glycosylation status affects the level of PC modification.

A) Figure shows immunoblots of whole cell lysates using the PC recognizing antibody TEPC-15 and the PilE peptide specific α-pilin antibody. The samples were loaded such that each lane showed equal numbers of cells. Strains used were wild-type (N400), pglE _ON (KS142), pglE _ON pptA (KS651), pilE _S63A pglE _ON (KS858), pilE _S63A pglE _ON pglC (KS859), pglC (KS649), pglC pptA (KS652), pglC pilT (KS860), 2xpilE (KS646), 2xpilE pglE _ON (KS655), 2xpilE pglE _ON pptA (KS656), 2xpilE _S63A pglE _ON (KS861), 2xpilE _S63A pglE _ON pglC (KS862), 2xpilE pglC (KS657), 2xpilE pglC pptA (KS658), 2xpilE pglC pilT (KS863), 3xpilE (KS647), 3xpilE pglE _ON (KS654), 3xpilE pglE _ON pptA (KS661), 3xpilE pglC (KS659), 3xpilE pglC pptA (KS660), and 3xpilE pglC pilT (KS864). All samples were run on the same gel and the dotted lines were introduced as guidance facilitating evaluation of the data. Figure B) shows immunoblots of whole cell lysates made from equal numbers of cells using the PC reactive antibody TEPC-15, the α-pilin antibody, and the trisaccharide (Ac-Gal₂-diNAcBac) specific monoclonal npg3 antibody. Strains used were pglE _ON pilF (KS851), pglE _ON pilF pptA (KS853), pglC pilF (KS852), and pglC pilF pptA (KS854). All samples on each blot were run on the same gel. Results representative of at least three different experiments are shown.

Figure 6. Influence of minor pilin-like proteins on PilE post-translational modifications.

Deconvoluted mass spectra of intact PilE from ESI MS analysis with multiple protein forms representing species differing in glycosylation and phospho-form modifications are shown. Minor peaks and peaks connected with Na⁺ and/or K⁺ adducts in the mass spectra are not marked. PilE was isolated from A) the wild-type (N400) background, B) the pilV mutant (KS790), C) the pilC mutant (KS789), D) the pilH mutant (KSKS801), E) the pilI mutant (KS804), F) the pilJ mutant (KS806), and G) from the pilK mutant (KS810).

Figure 7. MS analysis of PTMs in purified pili from a pilV pilT mutant.

A) Graphical representation of the relative abundance of phospho-form modified PilE compared to total PilE. B) Graphical representation of the relative abundance of various phospho-form- and glycan-modified PilE. The strain used was KS791 .

Figure 8. Identification of methylated PE on the PilE peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ in PilE.

A) MS spectrum between m/z 800-900 of the precursor masses corresponding to the quadruply charged peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ unmodified (at m/z 812.90) and mass additions consistent with PE (at m/z 843.65), monomethylated PE (mmPE) (at m/z 847.15), dimethylated PE (dmPE) (at m/z 850.65) and PC (at m/z 854.16) modification. Masses are reported as monoisotopic. The asterisk denotes unmodified peptide. B) The deconvoluted mass spectrum showing the monoisotopic and monoprotonated masses of the quadruple charged ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ peptide unmodified (at m/z 3246.56) and with mass additions consistent with PE (at m/z 3370.56), mmPE (at m/z 3384.58), dmPE (at m/z 3398.59) and PC (at m/z 3412.61) modification. The asterisk denotes unmodified peptide. C) MS2 HCD spectrum of the precursor peptide at m/z 843.65 [M+4H]⁴⁺ (observed monoisotopic mass of 3370.56 [M+H]⁺) confirming that peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ was modified with one PE. The reporter ion for PE at m/z 142.0 could be detected in the low mass area. D) MS2 HCD spectrum of the precursor peptide at m/z 847.15 [M+4H]⁴⁺ (observed monoisotopic mass of 3384.58 [M+H]⁺) confirming that peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ was modified with one mmPE. The reporter ion for mmPE at m/z 156.0 could be detected in the low mass area. E) MS2 HCD spectrum of the precursor peptide at m/z 850.65 [M+4H]⁴⁺ (observed monoisotopic mass of 3398.59 [M+H]⁺) confirming that peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ was modified with one dmPE. The reporter ion for dmPE at m/z 170.1 could be detected in the low mass area. F) MS2 HCD spectrum of the precursor peptide at m/z 854.16 [M+4H]⁴⁺ (observed monoisotopic mass of 3412.61 [M+H]⁺) confirming that peptide ⁵²SAVTEYYLNHGKWPENNTSAGVASPPTDIK⁸¹ was modified with one PC. The reporter ion for PC at m/z 184.1 could be detected in the low mass area. G) The structure of PE, mmPE, dmPE and PC together with their respective reporter ion m/z.