Fig 2: LEKTI expressing KLK5 KO OSCC cells display decreased malignant characteristics.A–L) Representative images showing that KLK5 KO OSCC cells partially inhibit wound closure (dashed white lines, wound edges). Inhibition of KLK5 with recombinant LEKTI treatment leads to a similar inhibition (dashed white lines, compare C, F, I and L with A, D, G and J) in wild-type cells. M) Quantification of the wound closure shows a delay, in KLK5 KO (red columns) and rLEKTI-treated KLK5 WT (gray columns) OSCC cells, when compared to the KLK5 WT OSCC cells (black columns), after 48 h and 72 h. Statistical analysis: two-way ANOVA and Tukey's Multiple Comparisons tests (p<0.02 = [*] and p<0.0001 = [***]). N) Quantification of BrdU positive cells using immunofluorescence shows that KLK5 KO OSCC cells (red column) have a decreased proliferation rate when comparing with the KLK5 WT OSCC cells (black column) following 1 h, 2 h and 3 h BrdU incubation. Statistical analysis: two-way ANOVA and Sidak's Multiple Comparisons tests, p = 0,0254 [*]; p = 0.0002 [***]; p<0,0001 [****]. O and P) KLK5 WT (black dashed circle) and KLK5 KO (red dashed circle) cells were xenografted, respectively, into the left and right flanks of nude mice (O, n = 8) and tumor growth was monitored every 3 days, starting at 40 days post injection, for 41 days. KLK5 WT (P, black) and KLK5 KO (P, red) tumors were followed for approximately 11 weeks when the endpoint was reached, and the mice were euthanized. Q and R) Images of dissected wild-type (Q, black asterisks) and KLK5 KO (Q, red asterisks) tumors. Necropsied tumors were weighed and plotted (R, KLK5 WT tumor, black dots; KLK5 KO tumor red dots). Statistical analysis: two-tailed paired t-test, p<0.005 [**]. S–W) Representative images of KLK5 WT and KLK5 KO tumors following anti-KLK5 and anti-LEKTI immunohistochemistry. As expected, KLK5 immunostaining is present only in the WT tumors (black arrows) LEKTI staining was more abundant in KLK5 WT tumors, however, it is present in both wild-type (T, black arrows) and KLK5 KO (W, black arrow) tumors. Negative controls were incubated in the absence of primary antibodies using KLK5 WT (U) and KLK5 KO (X) tumors.

Fig 3: Expression of LEKTI and AIRE in lymphatic tissue. AIRE-expressing mTEC cells in tonsil (a), LEKTI-expressing cells in the corresponding tonsillar area (b), AIRE-expressing cells around Hassall’s corpuscles in the thymus (c) and LEKTI expression in a corresponding thymic area (d). The original magnification in panels is: (a) 20x, insert 100x, (b) 20x, (c) 40x and (d) 40x

Fig 4: KLK5 expression decreases in OSCCs bearing higher levels of LEKTIA, D, G and J) Representative H&E staining of oral mucosa (A, O.M., n = 50), premalignant hyperkeratotic lesion (D, P.L., n = 50), well-differentiated OSCC (G, W.D.C., n = 122), and poorly differentiated OSCC (J, P.D.C., n = 79), bar: 25 µm. B, E, H and K) Representative IHC staining of LEKTI in O.M. (B), P.L. (E), W.D.C. (H), and P.D.C. (K), bar: 25 µm. C, F, I and L) Representative IHC staining of KLK5 in O.M. (C), P.L. (F), W.D.C. (I), and P.D.C. (L), bar: 25 µm. Black arrows indicate LEKTI-immunostained areas; Red arrows indicate KLK5-immunostained areas; Black arrowheads indicate dysplastic cells; Yellow arrowheads indicate nuclear polymorphism; Blue arrowheads indicate atypical mitosis; Asterisks indicate keratin pearls. M) Quantification of LEKTI expression in each experimental group,% of immunostained area. O.M., circles; P.L., squares; W.D.C., triangles; P.D.C., inverted triangles. Kruskal Wallis test followed by Dunn's Multiple Comparison test; p<0.0005[***]. N) Quantification of KLK5 expression in each experimental group,% of immunostained area. O.M., circles; P.L., squares; W.D.C., triangles; P.D.C., inverted triangles. Kruskal Wallis test followed by Dunn's Multiple Comparison test; p<0.05[*]; p<0.0005[***]. O) KLK5 protein expression in O.M (n = 50, circles), P.L. (n = 50, triangles) and O.S.C.C.s. samples (n = 201, diamonds) harboring low LEKTI (up to 1% of O.M. immunostaining, black circles, black triangles and black diamonds) and high LEKTI expression (>1% of O.M. immunostaining, gray circles, gray triangles and gray diamonds). Statistical analysis: ANOVA; p<0.05 [*]; p<0.005 [**].

Fig 5: Generation of CRISPR-driven KLK5 knockout OSCC cellsA) RT-PCR analysis of SPINK5 and KLK5 mRNA expression in HNSCC cell lines. Lanes 1–4 are Cal 27, UD-SCC-2, HN6 and HN30, respectively. No RNA was added in lane 5, and reverse transcriptase was omitted from the reactions in lane 6. B) Schematic diagram for the KLK5-specific guide RNA design. Blue arrows show guide RNAs 1 and 2 that target KLK5 Exon 2 (yellow box). Coding sequence is shown in purple and exons one and two are shown in gray and yellow boxes. C) Sequence analysis of viable KLK5 KO OSCC cell line (clone 6) exon 2 amplicon showing 11 nucleotide deletion (upper panel, red boxed sequence) and a premature stop codon (lower panel, red-boxed white asterisk). D) KLK5 western blot shows KLK5 protein band (red rectangle, lane 2) and the absence of KLK5 protein (red arrow, lane 3) in clone 6 lysate supernatant. Molecular weight standards and positive control using HEK293 cells overexpressing KLK5 are shown in lanes one and four, respectively. E) LEKTI western blot shows the 37-KDa LEKTI fragment in both KLK5 WT and KLK5 KO OSCC cell line (Clone 6) in vitro. Murine skin from mice overexpressing LEKTI in the epidermis (K5-LEKTI transgenic mouse) and murine skin from LEKTI-null mice were used, respectively, as positive and negative controls (8).