Proteomic Analysis and Discovery Using Affinity Proteomics and Mass Spectrometry*

Quantitative Proteomics Targeting Classes of Motif-containing Peptides Using Immunoaffinity-based Mass Spectrometry

The integrity of tight junctions is an important part of gastrointestinal health, as they are key to maintaining the intestinal barrier. In Ulcerative Colitis (UC), the intestinal barrier is compromised, leading to increased permeability for bacteria and other damaging components to cross the epithelial layer into the gastrointestinal tract, where it can cause an exacerbated inflammatory response and tissue damage. microRNAs are small non-coding RNAs that repress gene expression through the binding of mRNA. microRNA (miR)-24 has been studied in various cancers and is implicated in tumorigenesis through the suppression of various genes and proteins, including many relevant to tight junction integrity. We hypothesize that miR-24 is elevated in UC, where it acts to alter tight junction protein expression and causes barrier deficiency. We used qPCR to measure expression of miR-24 in human colon tissue and blood. We used the barrier-forming intestinal epithelial cell lines Caco-2 and T84 plated on transwells to study barrier integrity. We transfected these cells with miR-24 precursor in order to increase the expression of miR-24. Using transepithelial electrical resistance (TEER) and dextran flux experiments, we measured barrier integrity. Changes in gene expression and protein levels were measured via qPCR and western blot, respectively. miR-24 was elevated in human UC colon tissue compared to normal, cancer, Crohn’s disease, and Irritable Bowel Syndrome patient colon tissues. Furthermore, human UC patient blood had increased miR-24 compared to normal and Crohn’s Disease. miR-24 overexpression in Caco-2 and T84 cells caused a significant decrease in TEER and a significant increase in the amount of dextran flux through the cell monolayer, suggesting significant dysfunction of the barrier integrity. miR-24 overexpression also decreased apoptosis and had no effect on proliferation, suggesting that the increased paracellular permeability is a result of changes within the tight junction. This is supported by the differential regulation of various genes and proteins related to tight junctions, including cingulin, JAM-A, claudin-4, and occludin in miR-24 overexpressing cells. miR-24 is a significant regulator of tight junction integrity and its increased expression could play a role in intestinal barrier dysfunction in UC. As such, miR-24 is a novel therapeutic target to restore tight junction integrity in the gastrointestinal tract.

Colorectal cancer represents one of the most common neoplastic diseases worldwide, making it a frequent focus in routine pathological analyses. Visualizing complex three-dimensional (3D) structures, such as nerves within tumors, requires thick tissue sections, which necessitates the use of optical tissue-clearing methods to achieve transparency. However, following tissue clearing, samples typically require advanced imaging techniques such as light-sheet and two-photon confocal microscopy, which are usually unavailable in standard histological laboratories. We aimed to demonstrate how a well-established tissue-clearing approach can be adapted for use in a routine histological laboratory, enabling a robust 3D visualization of nerve fibers in samples of both normal human colon and colon cancer tissues. We modified the "clear unobstructed brain/body imaging cocktails" method, originally developed for whole-brain imaging in mice, and applied it to human colon tissue samples measuring approximately 10 mm3, a standard size typically processed in pathological laboratories. Our protocol, which integrates a tissue-clearing technique, enabled reliable immunofluorescent visualization of colonic nerve fibers labeled with anti-β3-tubulin antibodies. The labeled nerve fibers could be observed using a standard epifluorescence microscope, and high-quality 3D reconstructions were generated through a simple image analysis approach using the open-source software ilastik, which eliminates the need for confocal microscopy. The proposed steps provide a valuable method for researchers to visualize complex 3D structures, such as neural cells and processes, in both normal and tumor-transformed tissue settings.

Indian citrus ringspot virus (ICRSV) from a Kinnow mandarin orchard at the IARI, New Delhi, India (isolate K1) has flexuous particles 650 nm long, with clearly visible cross- banding. It can be mechanically transmitted to and multiplied in Phaseolus vulgaris cv. Saxa. The genome consists of one ssRNA of about 7.5 kb and the coat protein (CP) is 34 kd in size. The virus has been purified and an antiserum prepared suitable for DAS ELISA. Partial sequencing of the genome, revealing the CP ORF near the 3' end and further sequences upstream, indicates that ICRSV, while morphologically resembling a capillovirus, has no homology with this genus. Absence of serological relationship confirms this. Short amino acid motifs in the CP sequence reveal limited similarities with the genera Potexvirus, Carlavirus, Foveavirus and Allexivirus , but no strong similarity to any one of these, so that the virus does not fall into any known genus. RNA and cDNA probes that detect K1 are available, and also primers suitable for RT-PCR. EM decora- tion tests with an antiserum to K1 revealed the presence, in a Kagzi Kalan lime-lemon hybrid of a virus (KK) with similar morphology but differing consistently from K1 by one serological step. Moreover, in the original Kinnow orchard, we detected a further virus (K2), again with similar morphology, serologically related to K1 but differing by five steps. K2 now in turn appears, in pre- liminary results, to be a mixture containing yet a further morphologically similar virus, K3. Thus, Indian citrus ringspot disease may be caused by a series of viruses, and while diagnostic methods are now available for K1, these may not immediately solve the problem of field diagnosis.

The tyrosine kinase domain of the human insulin receptor (IR) contains several short amino acid motifs which are strictly conserved in all protein kinases and two sequence motifs which are specific to the tyrosine kinases (AAR or RAA and P(I)/VK/RWT/M). In the serine/threonine kinases these motifs are replaced by the sequences KPE and GT/SXXY/PX respectively. In the present work, the tyrosine kinase-specific sequences of the IR (1134AAR1136 and 1172PVRWM1176) were replaced using site-directed mutagenesis by sequences which confer a serine kinase specificity on the receptor. Five different IR mutants were expressed in Chinese hamster ovary (CHO) or COS cells and their structural and functional properties compared with those of the wild-type recombinant human IR. These mutants are processed normally and bind insulin with normal affinities. None of the mutants containing a putative serine kinase-specific sequence display detectable autophosphorylation or tyrosine kinase activity in response to insulin, either in vitro or in vivo. These mutants were also unable to phosphorylate serine/threonine kinase substrates after insulin stimulation. Unexpectedly, they showed impaired ATP binding, as studied by an original technique consisting of cross-linking adenosine 5'-([35S]thio)triphosphate to partially purified receptors. Finally, none of the studied mutants transmit the insulin signal necessary to stimulate either DNA or glycogen synthesis. These data provide evidence for the importance of these conserved sequences in the kinase domain for both receptor activation and kinase activity. Furthermore, they demonstrate that the exchange of sequences specific to the catalytic domain of tyrosine kinases for those specific to the serine/threonine kinases is not sufficient to confer serine/threonine specificity on the insulin receptor.

Two short amino acid motifs, WXGDXNXR and PXWCDRXL, define a large family of inositol polyphosphate 5-phosphatases. We tested the importance of seven of these conserved amino acids to substrate binding and catalysis by mutating each to alanine in the platelet 75 kDa inositol polyphosphate 5-phosphatase II (5-phosphatase II). Native and mutant forms of 5-phosphatase II were expressed in baculovirus-infected Sf9 cells, and the recombinant proteins were purified by Mono Q chromatography and studied for enzyme activity. Mutants D476A, N478A, D553A, and R554A had no detectable activity using all four known substrates for this enzyme. Mutants R480A, W551A, and I555A showed greatly reduced hydrolysis of Ins(1,4,5)P3 when compared to native enzyme [Km = 75 microM, Vm = 8300 nmol of Ins(1,4,5)P3 hydrolyzed min-1 (mg of protein)-1]. Mutants W551A and I555A had a Km for Ins(1,4,5)P3 hydrolysis similar to that of the native enzyme (35 microM and 81 microM, respectively), suggesting that these amino acids do not play a role in binding substrate. By contrast, mutant R480A had both increased Km (634 microM) and decreased Vm [855 nmol of Ins(1,4,5)P3 hydrolyzed min-1 (mg of protein)-1]. As judged by measurement of Km, mutant R480A retained normal binding of Ins(1,3,4,5)P4, suggesting that the arginine in motif 2 has a greater role in Ins(1,4,5)P3 binding than in Ins(1,3,4,5)P4 binding. Mutant I555A bound Ins(1,3,4,5)P4 with 8-fold reduced affinity. These mutations markedly reduced 5-phosphatase II hydrolysis of the three other substrates, Ins(1,3,4,5)P4, PtdIns(4,5)P2, and PtdIns(3,4,5)P3. We also tested a mutation comparable to D553A, D460A, in the 110 kDa form of the signaling inositol polyphosphate 5-phosphatase (5SIP110). 5SIP110 D460A had no detectable enzyme activity but retained the ability to bind GRB2. These results are consistent with a role for these conserved amino acids in substrate binding and catalysis.

Transcellular protein transduction using the Tat protein of HIV-1

RecD2 from Deinococcus radiodurans is a superfamily 1 DNA helicase that is homologous to the Escherichia coli RecD protein but functions outside the context of RecBCD enzyme. We report here on the kinetics of DNA unwinding by RecD2 under single and multiple turnover conditions. There is little unwinding of 20-bp substrates by preformed RecD2-dsDNA complexes when excess ssDNA is present to trap enzyme molecules not bound to the substrate. A shorter 12-bp substrate is unwound rapidly under single turnover conditions. The 12-bp unwinding reaction could be simulated with a mechanism in which the DNA is unwound in two kinetic steps with rate constant of k(unw) = 5.5 s(-1) and a dissociation step from partially unwound DNA of k(off) = 1.9 s(-1). These results indicate a kinetic step size of about 3-4 bp, unwinding rate of about 15-20 bp/s, and low processivity (p = 0.74). The reaction time courses with 20-bp substrates, determined under multiple turnover conditions, could be simulated with a four-step mechanism and rate constant values very similar to those for the 12-bp substrate. The results indicate that the faster unwinding of a DNA substrate with a forked end versus only a 5'-terminal single-stranded extension can be accounted for by a difference in the rate of enzyme binding to the DNA substrates. Analysis of reactions done with different RecD2 concentrations indicates that the enzyme forms an inactive dimer or other oligomer at high enzyme concentrations. RecD2 oligomers can be detected by glutaraldehyde cross-linking but not by size exclusion chromatography.

Insights into RNA Biology from an Atlas of Mammalian mRNA-Binding Proteins

The human colon synthetises several prostanoids which may have a role in inflammatory bowel diseases. As lipoxygenase products of arachidonate metabolism have been implicated in inflammatory processes, we have now investigated the formation of both lipoxygenase and cyclo-oxygenase metabolites from [14C]-arachidonic acid [(14C]-AA) by human colonic tissue. Homogenates of human colonic mucosa were incubated with [14C]-AA and after extraction into diethyl ether, separated by thin layer chromatography using two solvent systems that allowed resolution of cyclo-oxygenase and lipoxygenase products. The predominant cyclo-oxygenase products, as identified by their chromatographic mobility, were PGE2 greater than PGF2 alpha greater than PGD2 greater than TXB2 greater than 6-keto-PGF1 alpha. The formation of these products was inhibited both by indomethacin (1-10 microM) and the dual pathway inhibitor, BW755C (1-30 microM). The predominant lipoxygenase products formed, which had the chromatographic mobility of 11-, 12-, 15-HETE (which ran together) were inhibited by BW755C (19 microM) but not by indomethacin (3 microM). Further resolution of this TLC band, performed using normal phase HPLC, indicated that both 12-HETE and 15-HETE were major lipoxygenase products formed by human colonic homogenate. The present findings indicate that human colonic tissue can convert [14C]-AA into lipoxygenase as well as cyclo-oxygenase products and support the suggestion that lipoxygenase products may have a role in inflammatory bowel disease.

We developed two-photon (TP) probes for DNA (ABI-Nu), cytoplasm (Pyr-CT), and mitochondria (BF-MT). We found that ABI-Nu binds to AT in the minor groove, while ABI-Nu and BF-MT are effective for tracking in the cytoplasm and mitochondria, respectively. These probes showed very large effective two-photon action cross section values of 2230, 1555, and 790 Göppert-Mayer units (1 GM = 10−50 cm4 s photon−1molecule−1) at 740 nm with emission maxima at 473, 561, and 560 nm, respectively, in each organelle. Using these probes, we quantitatively estimated the mean nuclear area and the ratios of nuclei to cytoplasm and mitochondria to nuclei in human colon tissues by dual-colour two-photon microscopy imaging within 2 h after biopsy. The mean nuclear area and the nuclei to cytoplasm and mitochondria to cytoplasm ratios increased in the following order: normal colon mucosa <colon adenoma <colon adenocarcinoma. Furthermore, the nuclear areas of these tissues showed significant differences that were well outside of the ranges of experimental errors, indicating the diagnostic potential of this method.

Tissue-resident immune cells differ from their corresponding blood cells in many functional aspects. Although the proteome of blood immune cells has been well-investigated, there are almost no data on tissue-resident immune cells. Here, we explored the potential of using MALDI-TOF-MS imaging (MSI) to investigate these cells in colon tissue, which exhibits a strong infiltration of immune cells. MSI identified several proteinaceous markers that colocalized with specific structures of the colon, such as mucosa or muscularis mucosae, in six patients. In addition, we showed that certain m/z values have the same spatial distribution as CD3+ T lymphocytes in the lymphoid follicular structures or as CD206+ macrophages in the lamina propria. For further corroboration, blood lymphocytes and monocytes from 10 healthy volunteers were analyzed by intact cell mass spectrometry (ICMS). Furthermore, we analyzed monocyte-derived macrophages that had been polarized in vitro into proinflammatory M1 and anti-inflammatory M2 phenotypes. The mass spectra differed clearly among all immune cell types. Additionally, it was found that distinct signals from ICMS analysis were identical to the m/z values found in the MSI experiment in lymphoid follicular structures. These data show for the first time that MSI is well-suited to visualize the spatial distribution of immune cells in human colon tissue. We consider MALDI mass spectrometry imaging to be a technique with high potential for use in rapid investigations of tissue-specific features of cells.

Potentially valuable sources of DNA have been extracted from human colonic tissues and are retained in biobanks throughout the world, and might be re-examined to better understand host-microbe interactions in health and disease. However, the published protocols for DNA extraction typically used by gastroenterologists have not been systematically compared in terms of their recovery of the microbial fraction associated with colonic tissue. For this reason, we examined how three different tissue DNA extraction methods (the QIAGEN AllPrep DNA/RNA kit, salting out and high molecular weight (HMW) methods of DNA extraction) employed in past clinical trials, and the repeated bead beating and column (RBB+C) method might impact the recovery of microbial DNA from colonic tissue, using a custom designed phylogenetic microarray for gut bacteria and archaea. All four methods produced very similar profiles of the microbial diversity, but there were some differences in probe signal intensities, with the HMW method producing stronger probe intensities for a subset of the Firmicutes probes including Clostridium and Streptococcus spp. Real-time PCR analysis revealed that the HMW and RBB+C extracted DNA contained significantly more DNA of Firmicutes origin and that the different DNA extraction methods also gave variable results in terms of host DNA recovery. All of the methods tested recovered DNA from the archaeal community although there were some differences in probe signal intensity. Based on these findings, we conclude that while all four methods are efficacious at releasing microbial DNA from biopsy tissue samples, the HMW and RBB+C methods of DNA extraction may release more DNA from some of the Firmicutes bacteria associated with colonic tissue. Thus, DNA archived in biobanks could be suitable for retrospective profiling analyses, provided the caveats with respect to the DNA extraction method(s) used are taken into account.

The conversion of lysophosphatidic acid (LPA) to phosphatidic acid is carried out by the microsomal enzymes 1-acylglycerol-3-phosphate-O-acyltransferases (AGPATs). These enzymes are specific for acylating LPA at the sn-2 (carbon 2) position on the glycerol backbone and are important, because they provide substrates for the synthesis of phospholipids and triglycerides. At least, mutations in one isoform, AGPAT2, cause near complete loss of adipose tissue in humans. We cloned a cDNA predicted to be an AGPAT isoform, AGPAT11. This cDNA has been recently identified also as lysophosphatidylcholine acyltransferase 2 (LPCAT2) and lyso platelet-activating factor acetyltransferase. When AGPAT11/LPCAT2/lyso platelet-activating factor acetyltransferase cDNA was expressed in CHO and HeLa cells, the protein product localized to the endoplasmic reticulum. In vitro enzymatic activity using lysates of Human Embryonic Kidney-293 cells infected with recombinant AGPAT11/LPCAT2/lyso platelet-activating factor-acetyltransferase cDNA adenovirus show that the protein has an AGPAT activity but lacks glycerol-3-phosphate acyltransferase enzymatic activity. The AGPAT11 efficiently uses C18:1 LPA as acyl acceptor and C18:1 fatty acid as an acyl donor. Thus, it has similar substrate specificities for LPA and acyl-CoA as shown for AGPAT9 and 10. Expression of AGPAT11 mRNA was significantly upregulated in human breast, cervical, and colorectal cancer tissues, indicating its adjuvant role in the progression of these cancers. Our enzymatic assays strongly suggest that the cDNA previously identified as LPCAT2/lyso platelet-activating factor-acetyltransferase cDNA has AGPAT activity and thus we prefer to identify this clone as AGPAT11 as well.

Non-specific steroidal esterase activity and distribution in human and other mammalian tissues