Evidence of Alternative Cystatin C Signal Sequence Cleavage Which Is Influenced by the A25T Polymorphism

Abstract

Cystatin C (Cys C) is a small, potent, cysteine protease inhibitor. An Ala25Thr (A25T) polymorphism in Cys C has been associated with both macular degeneration and late-onset Alzheimer's disease. Previously, studies have suggested that this polymorphism may compromise the secretion of Cys C. Interestingly, we found that untagged A25T, A25T tagged C-terminally with FLAG, or A25T FLAG followed by green fluorescent protein (GFP), were all secreted as efficiently from immortalized human cells as their wild-type (WT) counterparts (e.g., 112%, 100%, and 88% of WT levels from HEK-293T cells, respectively). Supporting these observations, WT and A25T Cys C variants also showed similar intracellular steady state levels. Furthermore, A25T Cys C did not activate the unfolded protein response and followed the same canonical endoplasmic reticulum (ER)-Golgi trafficking pathway as WT Cys C. WT Cys C has been shown to undergo signal sequence cleavage between residues Gly26 and Ser27. While the A25T polymorphism did not affect Cys C secretion, we hypothesized that it may alter where the Cys C signal sequence is preferentially cleaved. Under normal conditions, WT and A25T Cys C have the same signal sequence cleavage site after Gly26 (referred to as 'site 2' cleavage). However, in particular circumstances when the residues around site 2 are modified (such as by the presence of an N-terminal FLAG tag immediately after Gly26, or by a Gly26Lys (G26K) mutation), A25T has a significantly higher likelihood than WT Cys C of alternative signal sequence cleavage after Ala20 ('site 1') or even earlier in the Cys C sequence. Overall, our results indicate that the A25T polymorphism does not cause a significant reduction in Cys C secretion, but instead predisposes the protein to be cleaved at an alternative signal sequence cleavage site if site 2 is hindered. Additional N-terminal amino acids resulting from alternative signal sequence cleavage may, in turn, affect the protease inhibition function of Cys C.

Fig 1. A25T Cys C is secreted as efficiently as WT Cys C from HEK-293T cells.

(A-C) HEK-293T cells were transfected with (A) untagged WT or A25T Cys C constructs, (B) WT or A25T Cys C FLAG constructs, or (C) WT or A25T Cys C FLAG GFP constructs. Forty-eight hours post transfection, conditioned media (40 μL) or cell lysates (40 μg) were run under denaturing, reducing conditions (with lithium dodecyl sulfate and DTT) on a 4–12% BOLT gel. Proteins were transferred to nitrocellulose membranes, probed with an anti-Cys C (1:500, Pierce), anti-FLAG M2 (1:2000, Sigma) or anti-β-actin (1:10,000, Sigma, or 1:2,000, LI-COR) antibody followed by an infrared-conjugated secondary antibody (1:10,000, LI-COR), representative data of at least 3 independent experiments. (D) Secreted protein bands were quantified using LI-COR software and data was presented as A25T Cys C secretion levels as percent of each respective WT Cys C control. n ≥ 7, mean ± S.D.

Fig 2. A25T Cys C is readily secreted from ARPE-19 cells.

(A) ARPE-19 cells were transfected with WT Cys C FLAG or A25T Cys C FLAG constructs. Forty-eight hours post transfection, cells were then treated with the indicated ER stressor (BFA = brefeldin A; Tg = thapsigargin; Tm = tunicamycin) or vehicle control (DMSO) for 24 h. 40 μL of conditioned media was then analyzed by western blotting using the anti-FLAG M2 antibody. (B) qPCR analysis of transfected ARPE-19 cells for changes in Cys C and unfolded protein response (UPR)-dependent transcripts (mean ± 95% C.I.). (C) Secretion of WT and A25T Cys C (DMSO treatment) was quantified by LI-COR and normalized for any differences in transcription based on the corresponding qPCR results (mean ± S.D.). (D) LI-COR quantification of secreted Cys C after drug treatments (mean ± S.D.). Representative data are shown for 3 independent experiments for A, and n ≥ 3 for panels B-D.

Fig 3. Signal sequence prediction of WT and A25T Cys C using SignalP 4.1 software.

(A, B) WT and A25T Cys C amino acid sequence was analyzed by SignalP 4.1 software and the resultant predicted cleavage sites, designated as site 1 and site 2, are shown. (C) In WT Cys C, site 1 cleavage occurs between Ala20 and Val21, site 2 cleavage occurs between Gly26 and Ser 27. The Ala residue in the A25T mutation is enlarged and colored green.

Fig 4. A G26K mutation eliminates Cys C signal sequence cleavage at site 2, and favors site 1 cleavage.

(A) WT, A25T, G26K or G26K/A25T Cys C FLAG constructs were transfected into HEK-293T cells and conditioned media samples were analyzed by western blotting. Representative data are shown for >3 independent experiments. (B) LI-COR quantification of secreted levels of Cys C variants listed in (A), n ≥ 3, mean ± S.D.).

Fig 5. The A25T polymorphism predisposes Cys C to alternative signal sequence cleavage.

(A) A FLAG tag after residue 20 (N20 FLAG) is predicted to eliminate site 2 cleavage. (B) However, a FLAG tag after residue 26 (N26 FLAG) is predicted to keep the two potential cleavage sites. (C) Secretion of N-terminal FLAG Cys C variants is similar. The secretion of N-terminal FLAG Cys C proteins (either N20 FLAG; FLAG tag after Ala20, or N26 FLAG; FLAG tag after Gly26) was assessed after plasmid transfection into HEK-293T cells. Representative data of at least 6 experiments. (D) Quantification of FLAG Cys C secretion in (C) by LI-COR software. n ≥ 6, mean ± S.D. (E) Differential IP of secreted FLAG Cys C from HEK-293T cells using either anti-FLAG M1 or anti-FLAG M2 beads. (F) Quantification of the relative levels of N-terminal (anti-FLAG M1 IP) to total (anti-FLAG M2 IP) FLAG-tagged protein. n ≥ 3, mean ± S.D., * = p < 0.05 by a paired t-test.

Fig 6. Mass spectrometry (MS) confirmation of Cys C signal sequence cleavage sites.

(A) ARPE-19 cells were infected with adenovirus encoding for the designated Cys C FLAG variants and the amount of secreted Cys C was isolated by IP using anti-FLAG M2 beads. Cys C was eluted using the FLAG peptide and run on an SDS-PAGE gel followed by Coomassie Blue staining. Asterisk denotes site 1 cleavage products, whereas the arrow denotes a newly identified cleavage product only found in A25T/G26K Cys C. (B) Intact MS of eluted Cys C purified from ARPE-19 conditioned media. Eluted, IP’d Cys C FLAG variants (not separated by SDS-PAGE) were analyzed by intact MS. Numbers above each peak denote molecular weight in Da. (C) Comparison of predicted and observed molecular weights and the corresponding predicted signal sequence cleavage site. (#) Note: the predicted molecular weight was based on assuming a loss of 4 Da due to two Cys C disulfide bonds and the loss of the C-terminal lysine (128 Da) during the purification process [55].