A divergent protein kinase A regulatory subunit essential for morphogenesis of the human pathogen Leishmania

Abstract

Parasitic protozoa of the genus Leishmania cycle between the phagolysosome of mammalian macrophages, where they reside as rounded intracellular amastigotes, and the midgut of female sand flies, which they colonize as elongated extracellular promastigotes. Previous studies indicated that protein kinase A (PKA) plays an important role in the initial steps of promastigote differentiation into amastigotes. Here, we describe a novel regulatory subunit of PKA (which we have named PKAR3) that is unique to Leishmania and most (but not all) other Kinetoplastidae. PKAR3 is localized to subpellicular microtubules (SPMT) in the cell cortex, where it recruits a specific catalytic subunit (PKAC3). Promastigotes of pkar3 or pkac3 null mutants lose their elongated shape and become rounded but remain flagellated. Truncation of an N-terminal formin homology (FH)-like domain of PKAR3 results in its detachment from the SPMT, also leading to rounded promastigotes. Thus, the tethering of PKAC3 via PKAR3 at the cell cortex is essential for maintenance of the elongated shape of promastigotes. This role of PKAR3 is reminiscent of PKARIβ and PKARIIβ binding to microtubules of mammalian neurons, which is essential for the elongation of dendrites and axons, respectively. Interestingly, PKAR3 binds nucleoside analogs, but not cAMP, with a high affinity similar to the PKAR1 isoform of Trypanosoma. We propose that these early-diverged protists have re-purposed PKA for a novel signaling pathway that spatiotemporally controls microtubule remodeling and cell shape.

Fig 1. PKAR3 is anchored to the DDM-insoluble subpellicular microtubules (SPMT) at the cortex of L. donovani cells.

(A) Schematic representation of selected Homo sapiens (Hsap_PKARIa), Saccharomyces cerevisiae (Scer_BCY1), Dictyostelium discoideum (Ddis_PKAR), Plasmodium falciparum (Pfal_PKAR) and Leishmania donovani (Ldon_PKAR1, Ldo_PKAR3) PKAR proteins. The scheme shows the dimerization and docking (D/D) domain (red), the RNI-like LRR domain (green), pseudo- inhibitor site (brown), and cyclic nucleotide binding (CNB) domains (blue). Orange and light blue indicate non-canonical D/D and CNB domains, respectively. (B) Indirect immunofluorescence using antibodies against PKAR3 (upper row), β-tubulin (middle row) and HSP83 (bottom row) in intact promastigote cells (left column) or the DDM-extracted cytoskeletons containing the SPMT (right column). Each picture is representative of at least 3 independent stainings. Note that for better visibility, the original scale bars were overpainted (C) Proteins from intact or DDM-extracted (supernatant and pellet) promastigotes were separated on 10% SDS-PAGE and the blotted proteins were subjected to western analysis using the antibodies indicated on the left side of the panels. The full-length gels are shown in S9 Fig.

Fig 2. PKAR3 binds to the SPMT via an N-terminal putative FH2 domain.

(A) Proteins extracted from intact and DDM-treated (supernatant and pellet) L. donovani promastigotes of the wild type (WT), PKAR3 null mutant (Δpkar3), as well as Δpkar3 ectopically expressing the full length PKAR3 (Δpkar3::FL) or the 90 aa truncated PKAR3 (Δpkar3::ΔN90). These proteins were separated by 10% SDS-PAGE and subjected to Western blot analysis using anti-PKAR3 (WT = 72 kDa, N90 = 70 kDa, upper row), anti-β-tubulin (middle row), and anti-HSP83 (lower row). The full-length gels are shown in S10A Fig. (B) Immunofluorescence of PKAR3 in Δpkar3::ΔN90 promastigotes using antibodies against PKAR3. Each picture is representative of at least 3 independent experiments. Note that for better visibility the original scale bars were overpainted. (C) Proteins extracted from intact or DDM-treated promastigotes of Δpkar3 expressing only the PKAR3 N-terminal 90 amino acids (Δpkar3::HA-N90). Proteins were analyzed by Western blotting using antibodies against the HA-tag (upper row), β-tubulin (middle row) and HSP83 (lower row). The full-length gels are shown in S10B Fig.

Fig 3. PKAR3 associates with tubulin of the SPMT.

(A) Promastigotes of L. donovani Δpkar3::FL labeled with rabbit anti-PKAR3 and mouse anti-β-tubulin, followed by secondary antibody detection with goat anti-rabbit antibodies conjugated to Alexa Fluor 568 and goat anti-mouse antibodies conjugated to Alexa Fluor 647. Fluorescence Resonance Energy Transfer (FRET) emission intensity was calculated relative to the darkness within the red borders. The levels of emission intensities are color coded (see rainbow ruler on the left) according to values obtained from the Axio Vision program (ranging between 0–0.7). (B) FRET analysis between PKAR3 and β-tubulin in Δpkar3 cells expressing Δpkar3::ΔN90 (Δpkar3::ΔN90), carried out as in panels A. (C) Comparing the number of FRET positive pixels per cell between Δpkar3::FL (panel A) and Δpkar3::ΔN90 cells (panel B). One-way ANOVA indicated that the number of FRET positive pixel containing Δpkar3::ΔN90/β-tubulin were 30% less than Δpkar3::FL/ β-tubulin (P<0.05, n = 3).

Fig 4. PKAR3 associates with PKAC3.

(A) Proteins from the DDM-soluble fractions of the T. brucei MiTat 1.2 PKAR1 null mutant (7) parental cell line (Δtbpkar1) and a line expressing a tetracycline-regulatable version of L. donovani PKAR3 containing a C-terminal Ty1-tag (LdR3-Ty1) were immunoprecipitated with mouse monoclonal antibodies against Ty1, followed by western blotting with rabbit antibodies against L. donovani PKAR3 and T. brucei PKAC3. The following abbreviations denote the different fractions interrogated: IN (input), FT (flow-through), and E (elution). The molecular weight marker is indicated in kilodaltons (kDa). An unidentified protein that cross-reacts with anti-PKAC3 antibody is indicated by an asterisk. (B) The same samples were probed with antibodies against PKAR3 and PKAC1. (C) L. donovani PKAR3 and PKAC3 containing N-terminal His₆ and Strep tags, respectively, were co-expressed in L. tarentolae and soluble proteins affinity purified on tandem Ni-NTA and Strep-Tactin columns. Aliquots from the input (IN) and eluate (E1, E2) fractions were analyzed by western blotting with rabbit antibodies against LdPKAR3, as well as mouse monoclonal antibodies against the His (Bio-Rad) and Strep (Qiagen) tags. (D) A similar experiment performed using tagged L. donovani PKAR1 and PKAC3 and probed with rabbit antibodies against TbPKAR1 that cross-react with LdPKAR1.

Fig 5. PKAC3 binding to the subpellicular microtubules is mediated by PKAR3.

(A) Wild type L. donovani promastigote (WT) and PKAR3 null mutant (Δpkar3) cell lines expressing HA-tagged PKAC3 were subjected to subpellicular microtubule enrichment using 0.5% DDM. Proteins extracted from these were subjected to western blot using antibodies against the HA-tag (upper row), PKAR3 (middle row) and β-tubulin (lower row). The full-length gels are shown in S11 Fig. (B) A gel slice containing proteins with a molecular mass range of 35–42 kDa was excised and subjected to mass spectrometry. The mean relative abundance of peptides from PKAC3, PKAR3, and actin (ACT) are shown for experiments using WT (blue bars) and Δpkar3 (orange bars) cell lines. Error bars show range of n = 2 independent repeats. (C) WT cell line expressing HA-tagged PKAC3 was labeled with rabbit anti-PKAR3 and mouse anti-HA, followed by goat anti-rabbit antibodies conjugated to Alexa fluor 568 and goat anti-mouse antibodies conjugated to Alexa fluor 647. The FRET emission intensity was calculated relative to the darkness within the red borders. Emission intensities are color coded (see rainbow ruler on the left) according to values obtained from the Axio Vision program, ranging between -0.01–0.1.

Fig 6. Structural modeling of the PKAR3/C3 complex and ligand binding.

(A) Protein sequence alignment of Bos taurus PKARIα (Btau_PKARIα) with L. donovani PKAR1 (Ldon_PKAR1) and PKAR3 (Ldon_PKAR3), indicating regions with known structural importance. Amino acids that are conserved in all three sequences are highlighted in black. (B) Structural modelling of the two nucleotide binding domains of Ldon_PKAR3 (pink) based on those from Btau_PKARIα (tan), highlighting the most important changes in the conserved amino acid sequence. (C) Binding isotherms from isothermal titration calorimetry (ITC) using refolded ligand-free recombinant protein containing the C-terminal portion (residues 321–647) of L. donovani PKAR3 titrated with cAMP. The graph gives the difference power (DP) between the reference and sample cells upon ligand injection as a function of time. The measurement corresponds to noise, excluding binding of cAMP. The results shown are from a representative example of three independent replicates. (D) Similar experiments performed with 7-CN-7-C-inosine (left) or toyocamycin (right). The upper panels show the difference power (DP) between the reference and sample cells after ligand injection as a function of time, while the lower panels show the total heat exchange per mole of injectant (integrated peak areas from upper panel) plotted against the molar ratio of ligand to protein. The K_D is the mean of three or more independent replicates, while the graphs are from a single representative replicate. (E) Structural model of the complex formed between Ldon_PKAR3 (pink) and Ldon_PKAC3 (blue) showing amino acids predicted to be involved in the salt-bridges and electrostatic interactions downstream of the pseudo-substrate. (F) Side chains of critical residues involved in the PKAR-PKAC interaction for experimentally determined Btau_PKARIα/Cα [PDBID:2QCS] (left) and predicted Ldon_PKAR3/C3 (right) complexes.

Fig 7. Deletion of LdPKAR3 or LdPKAC3 changes elongated L. donovani promastigote shape to rounded morphology.

(A) Microscopic DIC pictures of cells from late-log cultures of L. donovani promastigotes wild type (WT), PKAR3 null mutant (Δpkar3), PKAC3 null mutant (Δpkac3), and their add-back derivatives (Δpkar3::FL, Δpkar3::ΔN90, and Δpkac3::FL). (B) The percentage of elongated (black bars) or rounded (red bars) cells in the samples above as determined by Image Stream analysis. Ovoid cells (~30% in all cell lines) are not included in the graphs. Asterisks indicate statistically significant differences from WT of p <0.05 (*) or p< 0.01 (**) in three replicate experiments. (C) The mean flagellar length of wild type (left bar), Δpkar3 (middle bar) and Δpkac3 (right bar), with error bars indicating the range of three independent replicates.

Fig 8. Illustrated summary of conclusions (A-D) and proposed model of differentiation-induced morphogenesis (E).

(A) PKAR3 recruits PKAC3 to form a holoenzyme at the subpellicular microtubule cortex, thereby maintaining the elongated shape of WT promastigotes. In Δpkar3 or Δpkac3 null mutants (B and C, respectively), the PKAC3 no longer localizes to the SPMT, resulting in rounded promastigotes. Truncation of the putative FH2 domain at the N terminus of PKAR3 (D) also releases the PKAR3/C3 complex from the SPMT causing promastigote rounding. (E) Model of differentiation induced shape control incorporating published (indicated by asterisks) expression data for PKAC3. PKAC3 is expressed only in promastigotes *[8, 15, 17, 21]; the differentiation signal induces PKAC3 downregulation **[5]. PKA-dependent morphogenesis is only part of the induced differentiation program. Amastigote to elongated promastigote differentiation is reversible ***[7].