Dnase I Induces Other Endonucleases in Kidney Tubular Epithelial Cells by Its DNA-Degrading Activity

International Journal of Molecular Sciences

Article DNase I Induces Other Endonucleases in Kidney Tubular Epithelial Cells by Its DNA-Degrading Activity

Tariq Fahmi 1, Xiaoying Wang 1, Dmitry D. Zhdanov 1 , Intisar Islam 1, Eugene O. Apostolov 1, Alena V. Savenka 1 and Alexei G. Basnakian 1,2,*

1 Department of Pharmacology & Toxicology, University of Arkansas for Medical Sciences, 4301 West Markham Street, #638, Little Rock, AR 72205, USA; [email protected] (T.F.); [email protected] (X.W.); [email protected] (D.D.Z.); [email protected] (I.I.); [email protected] (E.O.A.); [email protected] (A.V.S.) 2 Central Arkansas Veterans Healthcare System, 4300 West 7th Street, Little Rock, AR 72205, USA * Correspondence: [email protected]; Tel.: +1-501-352-2870

Received: 16 September 2020; Accepted: 12 November 2020; Published: 17 November 2020

Abstract: Endonuclease-mediated DNA fragmentation is both an immediate cause and a result of apoptosis and of all other types of irreversible cell death after injury. It is produced by nine enzymes including DNase I, DNase 2, their homologs, caspase-activated DNase (CAD) and endonuclease G (EndoG). The endonucleases act simultaneously during cell death; however, regulatory links between these enzymes have not been established. We hypothesized that DNase I, the most abundant of endonucleases, may regulate other endonucleases. To test this hypothesis, rat kidney tubular epithelial NRK-52E cells were transfected with the DNase I gene or its inactive mutant in a pECFP expression vector, while control cells were transfected with the empty vector. mRNA expression of all nine endonucleases was studied using real-time RT-PCR; DNA strand breaks in endonuclease genes were determined by PCR and protein expression of the enzymes was measured by Western blotting and quantitative immunocytochemistry. Our data showed that DNase I, but not its inactive mutant, induces all other endonucleases at varying time periods after transfection, causes DNA breaks in endonuclease genes, and elevates protein expression of several endonucleases. This is the ﬁrst evidence that endonucleases seem to be induced by the DNA-degrading activity of DNase I.

Keywords: DNase I; endonucleases; DNA breaks; cell death; apoptosis

1. Introduction Apoptotic endonucleases are a group of endonucleases (DNases or DNase/RNases), which are important as DNA destroyers for cell death [1,2]. The term “apoptotic” is traditionally used, but is not exactly correct because these enzymes are active in all kinds of cell death [3]. Organs of the digestive system and kidney are known for their high activity of apoptotic endonucleases [4–7], where the enzymes are cytotoxic themselves or promote the cytotoxicity of other agents [1,8–10]. There are nine known apoptotic endonucleases [3], including deoxyribonuclease 1 (DNase I) [7,11], deoxyribonuclease 2 (DNase 2) [12], their homologs [13–15], caspase-activated DNase (CAD) [10,16], and endonuclease G (EndoG) [17–19]. Induction and nuclear import of the endonucleases is attributed to an irreversible stage of cell death manifested with the fragmentation of DNA [3,9,11]. Recent studies showed that inactivation of apoptotic endonucleases in the kidney or other organs prior to cell injury protects them against cell death, suggesting that DNA fragmentation is an important mediator of cell death [4,5,9,20]. Some of these apoptotic endonucleases have been known for many

Int. J. Mol. Sci. 2020, 21, 8665; doi:10.3390/ijms21228665 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2020, 21, 8665 2 of 16 years [21–23]; however, studying their specific roles in cell death is problematic for many reasons. For example, the same enzymes remain active in live cells and are responsible for DNA disposal after cell death [1,8,9], which makes it difficult to study the roles of these enzymes in live cells. Specific chemical inhibitors of these endonucleases suitable for in vivo use and are still under development. Genetic knockouts of these endonucleases often lack specific phenotypes [4,24–26]. All endonucleases catalyze the same reaction of destroying DNA by hydrolysis of phosphodiester bonds [1,8,9]. Endonuclease specificity to primary DNA structure is sometimes observed at the initial stages of DNA fragmentation, but it usually disappears quickly and is not present at later stages. All the enzymes share a role in cell death and DNA disposal which is present in all cells and tissues, and the endonucleases are often activated simultaneously in the same cell [1,27]. Other cell death enzymes, such as caspases and protein kinases, are known to act in networks by activating each other [28,29]. However, it remains to be established whether endonucleases are mechanistically linked with each other, or whether there is a network of endonucleases acting in a cooperative manner. Our previous study demonstrated that EndoG is capable of inducing other endonucleases [30]. It seemed logical to propose that DNase I, which is the most studied [4,5,7,11,20, 21,31], the most abundant [7], and the most active apoptotic endonuclease [4] would also induce other endonucleases. In this study, we evaluated whether the DNA-degrading activity of DNase I is the underlying mechanism of other endonucleases’ induction in kidney tubular epithelial cells.

2. Results

2.1. Generation of Inactive DNase I Mutant Inactive DNase I mutant, MT3cA, was produced as described in Materials and Methods. Kidney tubular epithelial cells were chosen because they are one of the primary sites of endonuclease expression in the body [4,32,33]. To determine whether the inactivation was complete, NRK-52E cells were transfected with a vector that encoded either native DNase I (DNase I) gene or mutated MT3cA DNase I, both of which were fused with CFP. Control cells were treated with an “empty” pECFP vector that encoded CFP protein. The total protein was extracted from transfected cells, and DNase I activity was assessed using an SRED assay. This method showed signiﬁcant DNase activity in cells transfected with pECFP-DNase I (Figure1). DNase I activity of cells transfected with pECFP-MT3cA was indistinguishable from activity in cells transfected with control pECFP indicating that the inactivation of all seven DNA binding sites in the DNase I molecule caused almost complete inactivation of its DNA-degrading activity.

2.2. Validation of DNase I or MT3cA Overexpression in Renal Tubular Epithelial Cells DNase I was previously shown to induce apoptosis after being overexpressed in mammalian cells [11]. Therefore, we first tested whether or not DNase I overexpression would induce cytotoxicity, which was determined by assessment of cell number, measurement of cell death and yields of total cellular DNA or RNA. NRK-52E cells were transfected with pECFP-DNase I, pECFP-MT3cA, or pECFP control plasmid. Based on image quantification, the transfection efficiency was not different between cells transfected with the three constructs in a period of time from 0 to 24 h after transfection (Figure1C,D). The analysis of the total DNA and RNA isolated from the transfected cells demonstrated a sufficient yield and an equally good quality of nucleic acids at every studied time point in a period of 24 h with no difference in respect to DNase I, MT3cA and control pECFP overexpression (data not shown). Real-time RT-PCR demonstrated no difference in the expression of DNase I in NRK-52E cells transfected with pECFP-DNase I or pECFP-MT3cA; however, both were significantly higher than DNase I expression in cells transfected with pECFP (Figure1E). The PI cell death assay showed a time-dependent cytotoxicity of all three constructs used for transfection. Overexpression of active DNase I tended to generate on average a higher cytotoxicity than pECFP or MT3cA, however the difference did not reach the level of statistical significance (Figure1F). These data indicate that at the Int. J. Mol. Sci. 2020, 21, 8665 3 of 16 used levels of overexpression, the cells were not significantly affected by the transfection to cause spontaneousInt. J. Mol. Sci. cell 2020 death., 21, x FOR PEER REVIEW 3 of 16

FigureFigure 1. Overexpression 1. Overexpression of of native native DNaseDNase I or or inacti inactiveve DNase DNase I mutant I mutant MT3cA MT3cA in kidney in kidney tubular tubular epithelialepithelial cells. cells. (A) The(A) DNaseThe DNase activity activity in the in total the proteintotal prot extractein extract of NRK-52E of NRK-52E cells 24cells h post 24 h transfection post wastransfection measured using was measured an SRED using assay. an Representative SRED assay. Representative images showing images DNA showing degradation DNA degradation circles in agar gel causedcircles in by agar pECFP-DNase gel caused by I, pECFP-MT3cApECFP-DNase I, orpECFP-MT3cA control pECFP or control transfection. pECFP (transfection.B) Quantification (B) of the SREDQuantification assay data of the (n SRED= 4, * assayp < 0.01 data compared (n = 4, * p to< 0.01 DNase compared I). (C )to Representative DNase I). (C) Representative fluorescent/phase fluorescent/phase contrast (CFP + PhC) images showing NRK-52E cells transfected with pECFP- contrast (CFP + PhC) images showing NRK-52E cells transfected with pECFP-DNase I, pECFP-MT3cA, DNase I, pECFP-MT3cA, or pECFP 8 h after transfection. (D) Total fluorescence of the transfected or pECFP 8 h after transfection. (D) Total fluorescence of the transfected cells 8 h after transfection. cells 8 h after transfection. (E) DNase I expression in NRK52E cells transfected with pECFP-DNase I, (E) DNasepECFP-MT3cA I expression (its inactive in NRK52E mutant) cells or transfectedpECFP measur withed by pECFP-DNase real-time RT-PCR I, pECFP-MT3cA 8 h after transfection. (its inactive mutant)mRNA or pECFPexpression measured levels were by real-timenormalized RT-PCR by 18 s 8 housekeeping h after transfection. gene (n = mRNA 4). (F) PI expression cell death levelsassay were normalizedshowing by the 18 mean s housekeeping percentage of gene dead (n =NRK-52E4). (F) PI cells cell transfected death assay with showing pECFP-DNase the mean I, percentagepECFP- of deadMT3cA, NRK-52E or pECFP cells transfected 0-24 h after with transfection pECFP-DNase (n = 4, I,* p pECFP-MT3cA, < 0.01 compared or to pECFP DNase 0-24 I). In h panel after transfectionF, the (n = difference4, * p < 0.01 in cell compared death was to not DNase statistically I). In panelsignificant (F), thebetween difference the constructs. in cell death was not statistically significant between the constructs. Int. J. Mol. Sci. 2020, 21, 8665 4 of 16

Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 4 of 16 2.3. DNase I Overexpression Induces Other Endonucleases 2.3. DNase I Overexpression Induces Other Endonucleases To determine whether the expression of any known apoptotic endonucleases is aﬀected by the overexpressionTo determine of DNase whether I, real-time the expression RT-PCR of was any performedknown apoptotic in both endonucleases pECFP-DNase is affected I-, pECFP-MT3cA- by the overexpression of DNase I, real-time RT-PCR was performed in both pECFP-DNase I-, pECFP- and pECFP-transfected cells every 4 h within 24 h after transfection (Figure2). DNase I expression was MT3cA- and pECFP-transfected cells every 4 h within 24 h after transfection (Figure 2). DNase I signiﬁcantlyexpression increased was significantly in pECFP-DNase increased I-transfected in pECFP-DNase cells comparedI-transfected to thosecells compared transfected to withthose pECFP at alltransfected time points. with Two pECFP of the at all studied time points. endonucleases, Two of the EndoGstudied endonucleases, and DNase γ, EndoG were induced and DNase by γ DNase, I as earlywere as induced 4 h post by transfection. DNase I as early DNase as 4 h X, post DNase transfection. I like 2, DNase DNase X, 2DNaseβ and I CADlike 2, wereDNase induced 2β and at 8 h. L-DNaseCAD 2 were and induced DNase at 2α 8 h.were L-DNase induced 2 and at DNase 12 h after2α were transfection. induced at 12 Therefore, h after transfection. all other Therefore, endonucleases wereall induced other endonucleases by DNase I overexpressionwere induced by withinDNase I 4 over to 24expression h post transfection. within 4 to 24 h post transfection.

pECFP-DNase I pECFP-MT3cA pECFP

DNase I DNase X DNase I like 2 4 * 12 * 20 * 10 3 * * 15 * 8 2 * 6 * 10 4 1 * * 5 2 * * expression 0 0 0

Normalized mRNA 0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 -1 -5 Time, hours Time, hours Time, hours

DNase Υ DNase 2α DNase 2β 20 * 4 4 * 15 3 3 10 * 2 2 * * 5 * 1 1

expression 0 0 0 0 4 8 12162024 Normalized mRNA mRNA Normalized 0 4 8 12 16 20 24 0 4 8 12162024 -5 -1 -1 Time, hours Time, hours Time, hours

L-DNase II CAD Endo G 12 * 3 5 * 10 2.5 4 * 2 * 8 3 6 1.5 2 4 * 1 * 1

expression 0.5 2 * * 0 0 0 Normalized mRNA mRNA Normalized -2 0 4 8 12162024 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Time, hours Time, hours Time, hours

FigureFigure 2. DNase 2. DNase I overexpression I overexpression causescauses induction of of other other endonucleases. endonucleases. NRK-52E NRK-52E cells were cells were transfectedtransfected with with pECFP, pECFP, pECFP-DNase pECFP-DNase I I oror pECFP-MT3cApECFP-MT3cA and and mRNA mRNA expressions expressions of endonucleases of endonucleases werewere measured measured by real-timeby real-time RT-PCR RT-PCR with with 44 h interv intervals.als. mRNA mRNA expression expression levels levels were werenormalized normalized by by 18 s housekeeping18 s housekeeping gene gene (n (=n 4,= 4, * *p p< < 0.050.05 compared to to pECFP pECFP and and pECFP-MT3cA). pECFP-MT3cA).

2.4. Overexpression2.4. Overexpression of Inactive of Inactive DNase DNase I I Mutant Mutant Does Not Not Induce Induce Other Other Endonucleases Endonucleases To determineTo determine whether whether the the inactivationinactivation of of DNA-de DNA-degradinggrading activity activity of DNase of DNase I would Ichange would or change or abolishabolish the the induction induction of other endonucleases, endonucleases, wewe compared compared the effect the e ﬀofect DNase of DNase I with Ithat with of thatits of its inactive mutant, MT3cA in the next experiment. A single time point of 8 h after transfection was inactive mutant, MT3cA in the next experiment. A single time point of 8 h after transfection was chosen chosen because at this time, as shown above, several endonucleases were induced by DNase I. Real- because at this time, as shown above, several endonucleases were induced by DNase I. Real-time

RT-PCR demonstrated a signiﬁcant decrease in the expression of all endonucleases in NRK-52E cells transfected with pECFP-MT3cA compared to pECFP-DNase I (Figure2). To evaluate the protein expression of individual endonucleases, total protein was extracted from NRK-52E cells 24 h after Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 5 of 16

time RT-PCR demonstrated a significant decrease in the expression of all endonucleases in NRK-52E Int. J. Mol. Sci. 2020, 21, 8665 5 of 16 cells transfected with pECFP-MT3cA compared to pECFP-DNase I (Figure 2). To evaluate the protein expression of individual endonucleases, total protein was extracted from NRK-52E cells 24 h after transfection with pECFP-DNase I or pECFP-MT3cA. Western blotting was performed with all transfection with pECFP-DNase I or pECFP-MT3cA. Western blotting was performed with all available available antibodies to endonucleases (Figure 3A). Quantification of the Western blotting data antibodies to endonucleases (Figure3A). Quantiﬁcation of the Western blotting data showed that showed that DNase I overexpression significantly increased expression of all endonucleases with the DNaseexception I overexpression of DNase I signiﬁcantly like 2 (Figure increased3B). Transfec expressiontion with pECFP-MT3cA of all endonucleases did not induce withthe expression exception of DNaseof I DNase like 2 (FigureX, DNase3B). 2β, TransfectionCAD, EndoG, and with DNase pECFP-MT3cA γ. Similarly to did pECFP-DNase not induce I, expression pECFP-MT3cA of DNase did X, DNasenot 2β ,change CAD, the EndoG, expression and DNase of DNaseγ. I Similarly like 2. Taken to pECFP-DNase together, these data I, pECFP-MT3cA indicate that the did inactivation not change the expressionof DNA-degrading of DNase I like activity 2. Taken of together, DNase I these abolishes data indicateits ability that to theinduce inactivation several ofother DNA-degrading cellular activityendonucleases. of DNase I abolishes In other words, its ability DNA to breaks induce are several the potential othercellular mechanism endonucleases. of other endonucleases’ In other words, DNA breaksinduction are by the DNase potential I. mechanism of other endonucleases’ induction by DNase I.

FigureFigure 3. Inactive 3. Inactive DNase DNase I mutant I mutant does does not not induce induce otherother endonucleases. endonucleases. (A) (WesternA) Western blotting blotting for for DNaseDNase X, DNase X, DNase I like I like 2, DNase 2, DNaseγ, γ DNase, DNase 2 β2β,, CAD,CAD, EndoG, and and GAPDH GAPDH (as (asa loading a loading control) control) in in NRK-52ENRK-52E cells 24cells h 24 after h after transfection transfection with with pECFP-DNase pECFP-DNase I Ior or pECFP-MT3cA. pECFP-MT3cA. (B) (QuantificationB) Quantiﬁcation of of DNase X, DNase I like 2, DNase γ, DNase 2β, CAD, and EndoG by densitometric analysis of band DNase X, DNase I like 2, DNase γ, DNase 2β, CAD, and EndoG by densitometric analysis of band mean mean optical densities (* p < 0.05 compared to pECFP-MT3cA; # p < 0.05 compared to pECFP- p p opticaltransfected densities NRK-52E (* < 0.05 cells). compared to pECFP-MT3cA; # < 0.05 compared to pECFP-transfected NRK-52E cells).

In another approach, to confirm that the protein expression of other endonucleases is not induced by MT3cA, we used quantitative immunocytochemistry to measure the protein expression of CAD and EndoG, two endonucleases which play the most active role in apoptotic DNA fragmentation and cell death [10,17,18]. The amount of CAD and EndoG proteins in cells transfected with pECFP-MT3cA was significantly less than in cells transfected with pECFP-DNase I (Figure4A,B). This additionally confirmed that the apoptotic endonucleases are induced by DNase I-generated DNA breaks. In parallel, we transfected another cell line ZR-75-1, which is known to have lower basal level of DNase I Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 6 of 16

In another approach, to confirm that the protein expression of other endonucleases is not induced by MT3cA, we used quantitative immunocytochemistry to measure the protein expression of CAD and EndoG, two endonucleases which play the most active role in apoptotic DNA fragmentation and cell death [10,17,18]. The amount of CAD and EndoG proteins in cells transfected Int. J. Mol.with Sci. pECFP-MT3cA2020, 21, 8665 was significantly less than in cells transfected with pECFP-DNase I (Figure6 of 16 4A,B). This additionally confirmed that the apoptotic endonucleases are induced by DNase I- generated DNA breaks. In parallel, we transfected another cell line ZR-75-1, which is known to have expressionlower basal [34] withlevel eitherof DNase pECFP-DNase I expression [34] I or wi pECFP-MT3cAth either pECFP-DNase followed I or by pECFP-MT3cA immunostaining followed for CAD and EndoG.by immunostaining Quantification for showedCAD and a significant EndoG. Quantification increase in the showed expression a significant of both CADincrease and in EndoG the in the cellsexpression transfected of both with CAD pECFP-DNase and EndoG in the I compared cells transfected to pECFP-MT3cA-transfected with pECFP-DNase I compared cells to (Figure pECFP-4C,D). AnotherMT3cA-transfected experiment was cells performed (Figure 4C,D). to determine Another whetherexperiment overexpression was performed of to EndoG determine instead whether of DNase overexpression of EndoG instead of DNase I would have a similar effect. NRK-52E cells were I would have a similar effect. NRK-52E cells were transfected with pECP-EndoG and pECFP (control) transfected with pECP-EndoG and pECFP (control) and immunostained for CAD. Quantitative and immunostainedimmunocytochemistry for CAD. showed Quantitative a significant immunocytochemistry increase in CAD expression showed in the cells a significant transfected increase with in CADEndoG expression compared in the to cells pECFP-transfected transfected with cells EndoG (Figure compared 4E). Therefore, to pECFP-transfected it seems that the nature cells of (Figure DNA 4E). Therefore,breaks itmay seems not that matter the naturefor all endonucleases of DNA breaks indu mayction: not they matter may for come all endonucleasesfrom one or another induction: they mayendonuclease. come from one or another endonuclease.

Figure 4. Inactive DNase I mutant does not induce EndoG or CAD proteins, while EndoG induces CAD. (A) CAD immunostaining and quantification of mean optical density (MOD) in NRK-52E cells 8 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.01 compared to pECFP-MT3cA). (B) EndoG immunostaining and quantification of MOD in NRK-52E cells 4 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.05 compared to pECFP-MT3cA). (C) CAD immunostaining and quantification of MOD in ZR-75-1 cells 8 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.01 compared to pECFP-MT3cA). (D) EndoG immunostaining and quantification of MOD in ZR-75-1 cells 4 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.05 compared to pECFP-MT3cA). (E) CAD immunostaining and quantification of MOD in NRK-52E cells transfected with pECFP-EndoG or pECFP 4 h after transfection (* p < 0.01 compared to pECFP). Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 7 of 16

Figure 4. Inactive DNase I mutant does not induce EndoG or CAD proteins, while EndoG induces CAD. (A) CAD immunostaining and quantification of mean optical density (MOD) in NRK-52E cells 8 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.01 compared to pECFP-MT3cA). (B) EndoG immunostaining and quantification of MOD in NRK-52E cells 4 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.05 compared to pECFP-MT3cA). (C) CAD immunostaining and quantification of MOD in ZR-75-1 cells 8 h after transfection with pECFP-DNase I or pECFP- Int. J. Mol. Sci.MT3cA2020 ,(*21 p, 8665< 0.01 compared to pECFP-MT3cA). (D) EndoG immunostaining and quantification of 7 of 16 MOD in ZR-75-1 cells 4 h after transfection with pECFP-DNase I or pECFP-MT3cA (* p < 0.05 compared to pECFP-MT3cA). (E) CAD immunostaining and quantification of MOD in NRK-52E cells 2.5. DNasetransfected I Overexpression with pECFP-EndoG InducesDegradation or pECFP 4 h after of Endonuclease transfection (* Genes p < 0.01 compared to pECFP).

We2.5. DNase then testedI Overexpression whether Induces DNase Degradation I overexpression of Endonuclease would Genes induce DNA breaks in endonuclease genes, and if this would coincide with the induction of endonuclease mRNAs. NRK-52E cells were We then tested whether DNase I overexpression would induce DNA breaks in endonuclease transfectedgenes, and with if pECFP-DNasethis would coincide I or with pECFP-MT3cA the induction and of endonuclease PCR was performed mRNAs. withNRK-52E DNA cells every were 4 h for up totransfected 24 h post with transfection. pECFP-DNase ThePCR I or pECFP-MT3cA for endonuclease and PCR genes was was performed performed with in DNA the every promoter 4 h for/exon 1 areas,up which to 24 h are post likely transfection to be involved. The PCR infor regulation endonuclease of thegenes genes. was performed To allow comparison,in the promoter/exon the PCR 1 was designedareas, towhich generate are likely products to be involved of approximately in regulation the of the same genes. length To allow or 3 kb.comparison, The results the PCR presented was in Figuredesigned5 show to some generate DNA products breaks (at of leastapproximately one double-stranded the same length or two or 3 single-stranded kb. The results presented DNA breaks in per 3 kb)Figure in the 5 promoter show some/exon DNA 1 regionbreaks (at in mostleast one of the double-stranded endonuclease or genes. two single-stranded This occurred DNA rather breaks late after transfection.per 3 kb) Inin particular,the promoter/exon the genes 1 region of DNase in most X, DNaseof the endonucleaseγ, DNase 2α genes., DNase This 2 βoccurred, L-DNase rather 2, and late CAD after transfection. In particular, the genes of DNase X, DNase γ, DNase 2α, DNase 2β, L-DNase 2, got these DNA breaks at varying time points between 12 and 24 h post transfection. The only two genes and CAD got these DNA breaks at varying time points between 12 and 24 h post transfection. The which showed no DNA breaks were DNase I like 2 and EndoG. Importantly, the degradation was only two genes which showed no DNA breaks were DNase I like 2 and EndoG. Importantly, the completelydegradation abolished was completely when MT3cA abolished mutant when was MT3cA used insteadmutant was of DNase used instead I. Therefore, of DNase the I. Therefore, degradation of endonucleasethe degradation genes of was endonuclease also dependent genes was on also the dependent DNA-degrading on the DNA-degrading activity of DNase activity I. of DNase I.

FigureFigure 5. DNase 5. DNase I overexpression I overexpression induces induces breaksbreaks in in th thee promoter/exon promoter/exon 1 regions 1 regions of endonucleases of endonucleases genes.genes. PCRs PCRs of genomic of genomic DNA DNA in in promoter promoter/exon/exon 1 regions regions of of endonucleases endonucleases and and housekeeping housekeeping genes genes (GAPDH,(GAPDH, PPL13a, PPL13a, and and Rplp0) Rplp0) were were performed performed usingusing primers li listedsted in in Table Table S1. S1.

2.6. Recombinant DNase I Induces Other Endonucleases To determine whether DNase I can directly induce the expression of other endonucleases by acting on chromatin, we used rat brain nuclei, which are notorious for their very low endogenous endonuclease activity [35]. The cell nuclei were isolated and exposed with varying concentrations of recombinant DNase I for 20 min. Total RNA was extracted and real-time RT-PCR was performed for EndoG and DNase γ mRNAs, the two endonucleases prominently induced by overexpression of DNase I as shown above. Both of the mRNAs were significantly induced at 0.1 ng/mL DNase I (Figure 6A,B). As expected, higher concentrations of DNase I reduced the expression of the endonucleases to the control levels or below, likely due to the excessive DNA degradation. To determine whether DNA breaksInt. J. Mol. of Sci.non-enzymatic2020, 21, 8665 origin would induce endonucleases in the same manner, we tested EndoG,8 of 16 DNase γ, and DNase I expression in isolated rat brain nuclei treated with varying concentrations of bleomycin,DNase γ, and a known DNase inducer I expression of oxidative in isolated DNA rat brea brainks nuclei[36]. Real-time treated with RT-PCR varying showed concentrations a significant of increasebleomycin, in the a known expression inducer of these of oxidative endonuclease DNA breakss at all [bleomycin36]. Real-time concentrations, RT-PCR showed which a signiﬁcantsuggested thatincrease non-enzymatic in the expression DNA ofbreaks these induce endonucleases the endonucl at alleases bleomycin similarly concentrations, to the enzymatic which breaks suggested (Figure that 6C,D).non-enzymatic DNA breaks induce the endonucleases similarly to the enzymatic breaks (Figure6C,D).

A B DNase γ EndoG 14 12 12 ** 10 * 10 8 8 6 6 expression

expression 4 4 Normalized mRNA

Normalized mRNA Normalized 2 2 0 0 0 10-5 10-4 10-3 10-2 10-1 010-5 10-4 10-3 10-2 10-1 recDNase I, µg/ml recDNase I, µg/ml C D EndoG DNase γ 6 ** 4 ** 5 * ** 3 * 4 * ** 3 2 * expression expression 2 1 Normalized mRNA

Normalized mRNA Normalized 1 0 0 0 1310 30 100 01310 30 100 Bleomycin, µg/ml Bleomycin, µg/ml E DNase I 5 ** 4 ** ** 3

2 * expression

Normalized mRNA 1

0 01310 30 100 Bleomycin, µg/ml

γ Figure 6. BothBoth DNase DNase I I and and bleomycin bleomycin induce induce EndoG EndoG and and DNase γ transcriptiontranscription in in isolated isolated rat rat brain brain A cellcell nuclei, while bleomycin also also induces induces DNase DNase I I transcription. transcription. InIn vitro vitro transcriptiontranscription of of EndoG EndoG ( (A)) and DNase γ (B) genes in isolated rat brain nuclei exposed with varying concentrations of recombinant and DNase γ (B) genes in isolated rat brain nuclei exposed with varying concentrations of DNase I (Pulmozyme) as measured by real-time RT-PCR (n = 4, * p < 0.05, ** p < 0.01 compared recombinant DNase I (Pulmozyme) as measured by real-time RT-PCR (n = 4, * p < 0.05, ** p < 0.01 to 0 concentration). (C–E) In vitro transcription measured by real-time RT-PCR of EndoG, DNase γ

and DNase I genes, respectively, in isolated rat brain nuclei exposed with varying concentrations of bleomycin. mRNA expression levels were normalized by 18 s housekeeping gene (n = 4, * p < 0.05, ** p < 0.01 compared to 0 concentration).

2.7. DNase I Digestion Results in DNA Unwinding To additionally test whether DNase I can induce expression of another endonuclease, supercoiled plasmid coding mouse EndoG and linearized plasmid coding mouse beta-actin (in vitro transcription kit control) were treated with recDNase I for up to 60 min followed by in vitro transcription and real-time RT-PCR. Beta-actin mRNA levels were high at 0 time point, meaning that the transcription Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 9 of 16

compared to 0 concentration). (C–E) In vitro transcription measured by real-time RT-PCR of EndoG, DNase γ and DNase I genes, respectively, in isolated rat brain nuclei exposed with varying concentrations of bleomycin. mRNA expression levels were normalized by 18 s housekeeping gene (n = 4, * p < 0.05, ** p < 0.01 compared to 0 concentration).

2.7. DNase I Digestion Results in DNA Unwinding To additionally test whether DNase I can induce expression of another endonuclease, supercoiledInt. J. Mol. Sci. plasmid2020, 21, 8665 coding mouse EndoG and linearized plasmid coding mouse beta-actin (in vitro9 of 16 transcription kit control) were treated with recDNase I for up to 60 min followed by in vitro transcription and real-time RT-PCR. Beta-actin mRNA levels were high at 0 time point, meaning that thekit transcription worked, and thenkit worked, went down and then to the went lowest down level to at the 60 lowest min (Figure level 7atB). 60 EndoG min (Figure mRNA 7B). expression EndoG mRNAwas initially expression low when was initially supercoiled low when DNA supercoiled was used, then DNA started was used, to increase then started after 10 to min, increase peaked after at 1020 min, min, peaked and the at decreased 20 min, and similar the decreased to beta-actin similar (Figure to 7beta-actinA). Digested (Figure plasmids 7A). Digested were subjected plasmids to weregel electrophoresis, subjected to gel whichelectrophoresis, showed how which digestion showed of how supercoiled digestion plasmid of supercoiled results inplasmid relaxation results of thein relaxationplasmid and of the its plasmid conversion and toits the conversion linear form. to the These linear data form. indicate These data that DNaseindicate I-mediated that DNase DNA I-mediated breaks DNAresult breaks in the result unwinding in the unwinding of DNA causing of DNA the causing increased the transcriptionincreased transcription of EndoG of gene. EndoG gene.

FigureFigure 7. 7. InIn vitro vitro transcriptiontranscription of of mouse mouse EndoG EndoG (A (A) )and and mouse mouse beta-actin beta-actin (B (B) )genes genes in in plasmid plasmid constructsconstructs cleaved cleaved with with recDNase recDNase I. I. mRNA mRNA expre expressionssion was was measured measured by by real-time real-time RT-PCR. RT-PCR. Arbitrary Arbitrary unitsunits reflect reﬂect coefficient coeﬃcient of of diluti dilutionsons from from 1:40 1:40 to to 1:320 1:320 that that were were applied applied to to the the standard standard curve curve (n (n == 4).4). (C)(C )Gel Gel electrophoresis electrophoresis shows shows digestion digestion of of pET29b-mEndoG pET29b-mEndoG and pTRI-β-actin plasmids byby recDNaserecDNase I I(Pulmozyme). (Pulmozyme).

2.8. Increased Expression of DNase I and Other Endonucleases, Except EndoG, Correlate with Cell Death The expressions of both DNase I and CAD measured by real-time RT-PCR highly correlated with cell death measured by PI assay. This is also true for other enzymes including DNase 2α, DNase I like 2, L-DNase II, DNase γ, DNase X, and DNase 2β. Surprisingly, EndoG was the only endonuclease whose expression did not correlate with cell death (Figure S1). EndoG is unique among the apoptotic endonucleases tested here in that it also has RNase activity which may aﬀect the role of EndoG in cell death apart from its DNase activity. Int. J. Mol. Sci. 2020, 21, 8665 10 of 16

3. Discussion This study is the first demonstration of a previously unknown ability of DNase I to induce other apoptotic endonucleases via DNA breaks. This group of endonucleases includes nine enzymes: four members of the DNase I family, three members of the DNase 2 family, CAD and EndoG [3,7,10–15,17,18]. The endonucleases can act independently by catalyzing the same reaction of hydrolysis of phosphodiester bonds immediately prior, during, and after cell death. However, regulatory links between the endonucleases have not previously been established. We hypothesized that DNase I, the most active apoptotic endonuclease, may regulate other endonucleases. To test this hypothesis, rat kidney tubular epithelial NRK-52E cells were transfected with rat Dnase I gene or its inactive mutant MT3cA in a pECFP expression vector, while control cells were transfected with the empty pECFP vector. We found that DNase I, but not its inactive mutant, induces all other endonucleases at varying time periods after transfection. DNase I overexpression also induced DNA breaks in the promoter/exon 1 regions of endonuclease genes, and protein expression of several of the endonucleases. In this transfection model, we showed that the entire group of endonucleases was regulated by the DNA-degrading activity of DNase I. Our experiments with isolated cell nuclei treated with recombinant DNase I suggest that the likely mechanism is a direct DNA unwinding by the DNase, making DNA accessible for transcription. Unexpectedly, we found that DNA degradation with the intensity of at least one double-stranded or two single-stranded DNA breaks per 3 kb induced by DNase I in the promoter/exon 1 regions of other endonuclease genes does not completely suppress RNA synthesis. Apparently, some endonuclease mRNAs continued to be synthesized at even higher rates after DNase I treatment despite the apparent start of DNA template destruction. This study is in agreement with our previous report that showed EndoG is able to induce other endonucleases [30]. It is also in agreement with the report from Yin et al. [5], which showed that the induction of EndoG expression in kidneys during cisplatin injury depends on the expression of DNase I. In another study, a DNase I inhibitor and the genetic inactivation of DNase I in mice regulated DNase γ endonuclease activity in the spleen [37,38]. After being simultaneously activated, DNase I and EndoG are likely to cooperate in the generation of double-stranded DNA cleavage products as described by Widlak et al. [39] DNase I was also shown to cooperate with DNase gamma and CAD [27]. Taken together, these data suggest the existence of a potential endonuclease network, which unite all apoptotic endonucleases as well as other enzymatic and non-enzymatic mechanisms of DNA fragmentation (Figure8). Since we observed DNase I acting through DNA breaks, we hypothesize that other DNA-breaking enzymes (for example, other apoptotic or DNA repair endonucleases or topoisomerases), reactive oxygen species (ROS), or physical and chemical agents (such as radiation or bleomycin) acting through DNA breaks would have the same effect of inducing apoptotic endonucleases. Therefore, DNase I is the upstream and downstream participant of the endonuclease network. It contributes to the pool of DNA breaks, and together with other endonucleases, seems to induced by the DNA breaks. There are two other enzyme networks known to associate with cell death, caspases and protein kinases [28,29]. There appears to be a common mechanism of self-promoting enzyme networks in cells which continue to work even during cellular destruction. The resiliency of this mechanism during events where cellular systems are being disintegrated highlights the importance of this enzyme-based cell-death system. Each of these networks is based on the activity of its members, proteinases or kinases, and thus is capable of acting independently of the normal regulatory transcriptional pathways. In this regard, the endonuclease network seems to work similarly. Future studies may need to determine whether other endonucleases would work similarly to DNase I, and if any and all non-enzymatic DNA breaks can induce endonucleases. It seems likely that the mediation of DNase I action through DNA breaks is a universal mechanism. Other endonucleases also can produce DNA breaks, or DNA breaks can be induced by physical and chemical agents. All of these would be expected to induce endonucleases. The important consequence of this finding may be Int. J. Mol. Sci. 2020, 21, 8665 11 of 16 that inhibiting a single endonuclease for tissue protection against injury is likely to be less efficient than inhibiting several endonucleases and other DNA-degrading activities at once. Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 11 of 16

DNase I Other DNA-damaging agents (ROS, radiation, bleomycin, DNA repair DNA strand breaks endonucleases, topoisomerases)

Chromatin decondensation, DNA unwinding

Increased transcription of endonuclease genes

Induction of apoptotic endonucleases (DNase X, DNase I like 2, DNase γ, DNase 2α, DNase 2β, L-DNase 2, CAD, EndoG)

FigureFigure 8.8. HypotheticalHypothetical schemescheme ofof apoptoticapoptotic endonucleaseendonuclease network.network.

OurThere data are indicate two other that enzyme DNase networks I acts as someknown kind to associate of a transcription with cell factor.death, Thiscaspases is similar and protein to the activationkinases [28,29]. of Fas transcriptionThere appears after to internalizationbe a common ofme extracellularchanism of DNaseself-promoting I reported enzyme by Oliveri networks et al. [40 in]. Althoughcells which DNase continue I DNA-cutting to work activityeven during is important cellular for destruction. induction of The other resiliency endonucleases, of this it mechanism is yet to be determinedduring events whether where DNase cellular I acts systems alone while are being fragmenting disintegrated DNA, or highlights if it employs the other importance endonucleases of this toenzyme-based do this as well. cell-death Current system. methods Each do of not these distinguish networks between is based DNA on the breaks activity induced of its bymembers, DNase Iproteinases and other endonucleases.or kinases, and Therefore,thus is capable in light of of acting our data, independently previously observedof the normal apoptotic regulatory DNA fragmentationtranscriptional induced pathways. by overexpression In this regard, ofthe DNase endonuclease I in COS cellsnetwork [11] seems can be explainedto work similarly. by activation of otherFuture endonucleases studies bymay DNase need I.to DNA determine breaks whether induced byother DNase endonucleases I and other endonucleaseswould work similarly will likely to induceDNase DNAI, and damage if any and pathways. all non-enzymatic Activation ofDNA endonucleases breaks can induce at different endonucleases. time points It after seems DNase likely I overexpression,that the mediation and theof existenceDNase I ofaction more through than one peakDNA (forbreaks example, is a foruniversal EndoG), mechanism. may indicate Other that moreendonucleases than one mechanism also can produce of endonuclease DNA breaks, induction or DNA links DNasebreaks Ican with be other induced endonucleases. by physical DNase and Ichemical is not the agents. most eAllffective of these instrument would be to expected degrade to DNA induce because endonucleases. it generates The single-stranded important consequence breaks, whichof this can finding be easily may repaired. be that inhibiting For the fastest a single degradation, endonuclease DNA for needs tissue to protec be cuttion bydouble-stranded against injury is DNAlikely breaks.to be less It wouldefficient be than interesting inhibiting to studysevera whetherl endonucleases DNase 2,and which other produces DNA-degrading double-stranded activities DNAat once. scissions, would induce endonucleases in the same way as DNase I. In addition, specificity of endonucleasesOur data indicate to induce that other DNase endonucleases I acts as some will kind require of a further transcription investigation. factor. This is similar to the activationIn summary, of Fas transcription we found that after DNase internalization I activity at of non-toxic extracellular levels DNase induced I reported the entire by Oliveri network et ofal. apoptotic[40]. Although/cytotoxic DNase endonucleases. I DNA-cutting The activity endonuclease is important network for induction seems to of be other an eff endonucleases,ective mechanism it is toyet ensure to be fastdetermined DNA degradation whether DNase and cellI acts death alone after while a deadly fragmenting stimulus. DNA, Multiple or if it endonucleasesemploys other workingendonucleases in a network to do this of expression as well. Current induction method woulds do provide not distinguish the flexibility between of responses DNA breaks and guarantee induced theby DNase effective I and DNA other fragmentation endonucleases. after Therefore, any stimulus. in light The of discovery our data, ofpreviously this endonuclease observed apoptotic network andDNA the fragmentation mechanism thatinduced mediates by over it mayexpression have broad of DNase implications. I in COS Itce provideslls [11] can a new be explained perspective by onactivation mechanisms of other of cell-death endonucleases which couldby DNase guide theI. DNA development breaks induced of cell-preserving, by DNase anti-apoptotic I and other therapyendonucleases for various will organlikely injuriesinduce DNA caused damage by trauma, pathways. hypoxia Activation/oxidant, of radiation, endonucleases drugs orat diseases.different Furthertime points studies after may DNase also I link overexpression, enzymatic DNA and fragmentation,the existence of DNA more repair, than one and peak non-enzymatic (for example, DNA for damageEndoG), in may a single indicate pathway. that more than one mechanism of endonuclease induction links DNase I with other endonucleases. DNase I is not the most effective instrument to degrade DNA because it generates single-stranded breaks, which can be easily repaired. For the fastest degradation, DNA needs to be cut by double-stranded DNA breaks. It would be interesting to study whether DNase 2, which produces double-stranded DNA scissions, would induce endonucleases in the same way as DNase I. In addition, specificity of endonucleases to induce other endonucleases will require further investigation.

Int. J. Mol. Sci. 2020, 21, 8665 12 of 16

4. Materials and Methods

4.1. DNase I and EndoG Cloning and Site-Direct Mutagenesis A previously cloned rat DNase I gene [41] or rat EndoG gene was inserted in the pECFP-N1 vector (Clontech, Mountain View, CA, USA) upstream of the ECFP gene. Inactive DNase I mutant, MT3cA, with seven DNA binding sites inactivated by mutation (R9G, R41A, Y76A, N170A, R111G, Y175A, Y211A), was produced using modiﬁed QuikChange® XL Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA).

4.2. Cell Culture Normal rat tubular epithelial NRK-52E cells (ATCC, Manassas, VA,USA) were grown in Dulbecco’s Modified Eagle’s Medium (DMEM) (ATCC) supplemented with 5% fetal bovine sera. Human breast cancer ZR-75-1 cells (ATCC) were grown RPMI-1640 with 10% fetal bovine sera. Cells were cultured at 5% CO2/95% air in a humidified atmosphere at 37 ◦C, fed at intervals of 48–72 h, and used within 1 day after confluence. The cells were transfected using Lipofectamine 3000 (Invitrogen, Grand Island, NY, USA) and the expression of fusion DNase I-cyan fluorescent protein (CFP) or CFP alone (control) was detected by fluorescent microscopy.

4.3. Propidium Iodide Cell Death Assay

Transfected NRK-52E cells were incubated at 37 ◦C with 40 µM propidium iodide (PI) for 30 min, and the ﬂuorescence of dead cells was measured in a plate reader at 530/645 nm (reading R1). Cells were then permeabilized with 10% Triton X100 for 20 min, and the total ﬂuorescence was measured at the same wavelength (reading R2). The percentage of dead cells was calculated as the ratio R1/R2.

4.4. Isolation of Cell Nuclei and In Vitro Transcription Rat brain nuclei were isolated as previously described [35]. The nuclei were exposed with varying concentrations of human recombinant DNase I (Pulmozyme, Genentech, San Francisco, CA, USA) or bleomycin (Sigma, St. Louis, MO, USA) for 20 min in 50 mM Tris-HCl, pH 7.9, 0.25 M sucrose, 10 mM MgCl2, 1 mM CaCl2, 5 mM 2-mercaptoethanol. Total RNA was extracted and real-time RT-PCR was performed as described below. For in-vitro transcription, Mouse-EndoG/pET29b (Novagen) and linear pTRI-β-actin-Mouse contained T7 promoter (Ambion, Austin, TX, USA) were exposed to human recombinant DNase I (recDNase I) (0.1 µg/mL) (Pulmozyme, Genentech, San Francisco, CA, USA) for varying time periods, and then Maxi script T7 kit (Ambion, Austin, TX, USA) was used according to the manufacturer protocol.

4.5. Single Radial Enzyme Diﬀusion (SRED) Assay Total protein extracts were prepared from NRK-52E cells as previously described [5]. Protein was measured using Bradford protein assay (Pierce, Rockford, IL, USA). Bovine serum albumin was used as the standard. DNase activity was determined using SRED assay in total protein extracts [42]. For that, Difco DNase Test Agar (BD, Sparks, MD, USA) was reconstituted in water (4.2% w/v), autoclaved, and distributed to 100 mm Petri dishes. Series of 3 mm holes were prepared in cold agar and 5 µL of cell lysate were applied to each hole for 16 h at 37 ◦C. Remaining DNA was visualized by precipitation with 1M HCl followed by staining with ethidium bromide, and imaged with the Eagle Eye imaging system. Human recombinant DNase I (Genentech) was used as the positive control. Densitometric analyses were done by using Image J 1.30v, in which the mean optical density was calculated per a ﬁxed area in the circle around each well. Int. J. Mol. Sci. 2020, 21, 8665 13 of 16

4.6. RNA and DNA Extraction Total RNA from NRK-52E cells or rat brain nuclei was extracted using RNeasy Mini Kit (QIAGEN, MD, USA) followed by DNA removal using RNase-free DNase kit (QIAGEN GmbH, Hilden, Germany). DNA was extracted from cultured NRK-52E cells using DNeasy blood and tissue kit (QIAGEN).

4.7. PCR and Real-Time RT-PCR PCR was done using extracted DNA from the NRK52E cells in a 25-µL reaction using SmartCycler (Cepheid, Sunnyvale, CA, USA). Reaction mix was prepared using Advantage®-GC2 PCR kit (Clontech, Mountain View, CA, USA) according to manufacturer recommendations. All primers for the endonucleases and their annealing temperatures are shown in Table S1. PCR products were separated in 1% agarose gel, stained with ethidium bromide and photographed under UV light in BIO-RAD imaging system (Hercules, CA, USA). For real time RT-PCR, the previously described protocol was followed [34]. Briefly, total RNA (5 µg) was reverse-transcribed in a 50-µL reaction followed by real-time RT-PCR in a 25-µL reaction using SmartCycler (Cepheid). Reaction mix was prepared using Platinum SYBR Green qPCR Supermix-UDG (Invitrogen) according to manufacturer recommendations. Two-temperature cycles with annealing/extension were used. The fluorescence was measured at the end of the annealing step. The melting curve analyses were performed at the end of the reaction (after 45th cycle) between 60 ◦C and 95 ◦C to assess the quality of the final PCR products. The threshold cycles, C(t), values were calculated by fixing the basal fluorescence at 15 units. The standard curve of the reaction effectiveness was performed using the serially diluted (5 points) mixture of all experimental cDNA samples in triplicates for each endonuclease and 18 s RNA separately. Calculation of the relative RNA concentration was performed using Cepheid SmartCycle software (Version 2.0d). Data are presented as the ratio of endonuclease/18 s mRNA.

4.8. Western Blotting Protein was separated in an 11.5% gel according to the Laemmli’s procedure [43]. The total protein extract from cells (10–100 µg) was dissolved in 50 mM Tris-HCl, pH 6.8, 1% SDS, 2 mM EDTA, 1% 2-mercaptoethanol and 7.5% glycerol, and denatured by heating at 100 ◦C for 10 min. Electrophoresis was performed at 100 V for 2 h. Proteins were transferred onto the nitrocellulose membrane in Novex transferring buﬀer (Invitrogen) at 40 V for 3 h. The membranes were stained with Ponseau S (Sigma, St. Louis, MO, USA) to control the equal protein load as described elsewhere [44]. After soaking in the blocking solution overnight at 4 ◦C, the membrane was incubated with polyclonal anti-EndoG antibody (Millipore, Billerica, MA, USA) diluted 1:1000, polyclonal Anti-CAD antibody (Abbiotec, San Diego, CA, USA) diluted 1:500, polyclonal anti-GAPDH (Abcam, Cambridge, MA, USA) antibody diluted 1:1000, or anti-DNase γ, anti-DNase 2β, anti-DNase X and DNase I like 2 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA) diluted 1:500, then washed in Tris-buﬀered saline (TBS), pH 7.6. Primary antibodies, including anti-EndoG, anti-CAD and anti-GAPDH were detected with anti-rabbit IgG-horseradish peroxidase (HRP) while anti-DNase γ, anti-DNase 2β, anti-DNase X and DNase I like 2 were detected with donkey anti-goat IgG-horseradish peroxidase (HRP) using SuperSignal chemiluminescent kit (Pierce Biotechnology, Rockford, IL, USA).

4.9. Immunocytochemistry and Image Analyses Cells were fixed with 5% formalin and the immunostaining was performed as described previously [45]. The cells were immunostained with rabbit anti-EndoG at 1:200 dilution (Millipore) or anti-CAD at 1:100 dilution (Abbiotec) in dilution buffer (0.5% BSA, 0.05% Tween-20, phosphate saline buffered (PBS)). The primary antibodies were detected with 1:400 diluted anti-rabbit IgG-AlexaFluor 594 conjugates (Invitrogen). Control sections were performed by substituting the primary antibody with dilution buffer only. Cells were mounted under cover slips with Prolong® Antifade kit (Invitrogen, Carlsbad, CA, USA) and acquired using the Olympus IX-51 inverted microscope (Olympus America, Int. J. Mol. Sci. 2020, 21, 8665 14 of 16

Center Valley, PA, USA) equipped with an ORCA-ER monochrome camera (Hamamatsu Photonics K.K., Hamamatsu City, Japan). CFP and AlexaFluor 594 were visualized using the following settings for excitation/emission: 438/483 and 562/624 respectively. Image analysis was performed using SlideBook 6.2 software. For quantification, 10 independent fields of view were collected per each rat kidney section, and mean optical density (MOD) was recorded for AlexaFluor 594 channel. The data were presented as averages of MODx/field of view for each channel.

4.10. Statistics Statistical analysis was performed with 2-way ANOVA and Student’s t-test. Results were expressed as mean standard error of mean (SEM). p < 0.05 was considered significant. ± Supplementary Materials: Supplementary Materials can be found at http://www.mdpi.com/1422-0067/21/22/ 8665/s1. Author Contributions: Conceptualization, A.G.B.; methodology, A.G.B. and T.F.; experiments and data analysis, T.F., X.W., D.D.Z., I.I., E.O.A. and A.V.S.; writing—original draft preparation, T.F. and D.D.Z.; writing—review and editing, E.O.A. and A.G.B.; supervision, A.G.B.; funding acquisition, A.G.B. All authors have read and agreed to the published version of the manuscript. Funding: This study was supported by National Institutes of Health grant 2P20 GM109005-06 and VA Merit Review grant 2I01 BX002425. Acknowledgments: We acknowledge the DNA Damage and Toxicology Core of the University of Arkansas for Medical Sciences for performing the immunohistochemistry staining experiments. We thank Pamela J. Kahler and Christopher L. Moore for editorial assistance. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Hengartner, M.O. Apoptosis. DNA destroyers. Nature 2001, 412, 2729. [CrossRef] 2. Majtnerova, P.; Rousar, T. An overview of apoptosis assays detecting DNA fragmentation. Mol. Biol. Rep. 2018, 45, 1469–1478. [CrossRef][PubMed] 3. Basnakian, A.G.; Braman, N.; Yin, X.; Kolomeychuk, S.; Apostolov, E.O. Toxicological Analysis Using DNA Fragmentation. In Encyclopedia of Analytical Chemistry; Meyers, R.A., Ed.; John Wiley: Chichester, UK, 2017; pp. 1–12. [CrossRef] 4. Basnakian, A.G.; Apostolov, E.O.; Yin, X.; Napirei, M.; Mannherz, H.G.; Shah, S.V. Cisplatin nephrotoxicity is mediated by deoxyribonuclease I. J. Am. Soc. Nephrol. 2005, 16, 697–702. [CrossRef][PubMed] 5. Yin, X.; Apostolov, E.O.; Shah, S.V.; Wang, X.; Bogdanov, K.V.; Buzder, T.; Stewart, A.G.; Basnakian, A.G. Induction of Renal Endonuclease G by Cisplatin Is Reduced in DNase I-Deficient Mice. J. Am. Soc. Nephrol. 2007, 18, 2544–2553. [CrossRef][PubMed] 6. Singh, M.; Odeniyi, D.T.; Apostolov, E.O.; Savenka, A.; Fite, T.; Wangila, G.W.; Walker, R.B.; Basnakian, A.G. Protective effect of zinc-N-acetylcysteine on the rat kidney during cold storage. Am. J. Physiol. Renal. Physiol. 2013, 305, F1022–F1030. [CrossRef] 7. Lacks, S.A. Deoxyribonuclease I in mammalian tissues. Specificity of inhibition by actin. J. Biol. Chem. 1981, 256, 2644–2648. 8. Samejima, K.; Earnshaw, W.C. Trashing the genome: The role of nucleases during apoptosis. Nat. Rev. Mol. Cell Biol. 2005, 6, 677–688. [CrossRef] 9. Apostolov, E.O.; Wang, X.; Shah, S.V.; Basnakian, A.G. Role of EndoG in development and cell injury. Cell Death Differ. 2007, 14, 1971–1974. [CrossRef] 10. Nagata, S. Apoptotic DNA fragmentation. Exp. Cell Res. 2000, 256, 12–18. [CrossRef] 11. Polzar, B.; Peitsch, M.C.; Loos, R.; Tschopp, J.; Mannherz, H.G. Overexpression of deoxyribonuclease I (DNase I) transfected into COS-cells: Its distribution during apoptotic cell death. Eur. J. Cell Biol. 1993, 62, 397–405. 12. Evans, C.J.; Aguilera, R.J. DNase II: Genes, enzymes and function. Gene 2003, 322, 1–15. [CrossRef][PubMed] 13. Shiokawa, D.; Hatanaka, T.; Hatanaka, M.; Shika, Y.; Nishikawa, A.; Tanuma, S. cDNA cloning and functional characterization of Xenopus laevis DNase gamma. Apoptosis 2006, 11, 555–562. [CrossRef][PubMed] Int. J. Mol. Sci. 2020, 21, 8665 15 of 16

14. Shiokawa, D.; Matsushita, T.; Shika, Y.; Shimizu, M.; Maeda, M.; Tanuma, S. DNase X is a glycosylphosphatidylinositol-anchored membrane enzyme that provides a barrier to endocytosis-mediated transfer of a foreign gene. J. Biol. Chem. 2007, 282, 17132–17140. [CrossRef][PubMed] 15. Napirei, M.; Wulf, S.; Eulitz, D.; Mannherz, H.G.; Kloeckl, T. Comparative characterization of rat deoxyribonuclease 1 (Dnase1) and murine deoxyribonuclease 1-like 3 (Dnase1l3). Biochem. J. 2005, 389, 355–364. [CrossRef][PubMed] 16. Larsen, B.D.; Sorensen, C.S. The caspase-activated DNase: Apoptosis and beyond. FEBS J. 2017, 284, 1160–1170. [CrossRef] 17. Parrish, J.; Li, L.; Klotz, K.; Ledwich, D.; Wang, X.; Xue, D. Mitochondrial endonuclease G is important for apoptosis in C. elegans. Nature 2001, 412, 90–94. [CrossRef] 18. Widlak, P.; Garrard, W.T. Discovery, regulation, and action of the major apoptotic nucleases DFF40/CAD and endonuclease G. J. Cell Biochem. 2005, 94, 1078–1087. [CrossRef] 19. Li, S.; Wang, T.; Zhai, L.; Ge, K.; Zhao, J.; Cong, W.; Guo, Y. Picroside II Exerts a Neuroprotective Effect by Inhibiting mPTP Permeability and EndoG Release after Cerebral Ischemia/Reperfusion Injury in Rats. J. Mol. Neurosci. 2018, 64, 144–155. [CrossRef] 20. Napirei, M.; Basnakian, A.G.; Apostolov, E.O.; Mannherz, H.G. Deoxyribonuclease 1 aggravates acetaminophen-induced liver necrosis in male CD-1 mice. Hepatology 2006, 43, 297–305. [CrossRef] 21. Kunitz, M. Crystalline desoxyribonuclease; digestion of thymus nucleic acid; the kinetics of the reaction. J. Gen. Physiol. 1950, 33, 363–377. [CrossRef] 22. Koizumi, T. Deoxyribonuclease II (DNase II) activity in mouse tissues and body fluids. Exp. Anim. 1995, 44, 169–171. [CrossRef][PubMed] 23. Ruiz-Carrillo, A.; Renaud, J. Endonuclease G: A (dG)n X (dC)n-specific DNase from higher eukaryotes. EMBO J. 1987, 6, 401–407. [CrossRef][PubMed] 24. Irvine, R.A.; Adachi, N.; Shibata, D.K.; Cassell, G.D.; Yu, K.; Karanjawala, Z.E.; Hsieh, C.L.; Lieber, M.R. Generation and characterization of endonuclease G null mice. Mol. Cell Biol. 2005, 25, 294–302. [CrossRef] 25. David, K.K.; Sasaki, M.; Yu, S.W.; Dawson, T.M.; Dawson, V.L. EndoG is dispensable in embryogenesis and apoptosis. Cell Death Differ. 2006, 13, 1147–1155. [CrossRef][PubMed] 26. Los, M.; Neubuser, D.; Coy, J.F.; Mozoluk, M.; Poustka, A.; Schulze-Osthoff, K. Functional characterization of DNase X, a novel endonuclease expressed in muscle cells. Biochemistry 2000, 39, 7365–7373. [CrossRef] 27. Koyama, R.; Arai, T.; Kijima, M.; Sato, S.; Miura, S.; Yuasa, M.; Kitamura, D.; Mizuta, R. DNase gamma, DNase I and caspase-activated DNase cooperate to degrade dead cells. Genes Cells Devoted Mol. Cell. Mech. 2016, 21, 1150–1163. [CrossRef] 28. Li, J.; Yuan, J. Caspases in apoptosis and beyond. Oncogene 2008, 27, 6194–6206. [CrossRef] 29. Yahiro, K.; Wada, A.; Yamasaki, E.; Nakayama, M.; Nishi, Y.; Hisatsune, J.; Morinaga, N.; Sap, J.; Noda, M.; Moss, J.; et al. Essential domain of receptor tyrosine phosphatase beta (RPTPbeta) for interaction with Helicobacter pylori vacuolating cytotoxin. J. Biol. Chem. 2004, 279, 51013–51021. [CrossRef] 30. Zhdanov, D.D.; Fahmi, T.; Wang, X.; Apostolov, E.O.; Sokolov, N.N.; Javadov, S.; Basnakian, A.G. Regulation of Apoptotic Endonucleases by EndoG. DNA Cell Biol. 2015, 34, 316–326. [CrossRef] 31. Napirei, M.; Karsunky, H.; Zevnik, B.; Stephan, H.; Mannherz, H.G.; Moroy, T. Features of systemic lupus erythematosus in Dnase1-deficient mice. Nat. Genet. 2000, 25, 177–181. [CrossRef] 32. Basnakian, A.G.; Ueda, N.; Kaushal, G.P.; Mikhailova, M.V.; Shah, S.V. DNase I-like endonuclease in rat kidney cortex that is activated during ischemia/reperfusion injury. J. Am. Soc. Nephrol. 2002, 13, 1000–1007. [PubMed] 33. Kaushal, G.P.; Basnakian, A.G.; Shah, S.V. Apoptotic pathways in ischemic acute renal failure. Kidney Int. 2004, 66, 500–506. [CrossRef][PubMed] 34. Basnakian, A.G.; Apostolov, E.O.; Yin, X.; Abiri, S.O.; Stewart, A.G.; Singh, A.B.; Shah, S.V. Endonuclease G promotes cell death of non-invasive human breast cancer cells. Exp. Cell Res. 2006, 312, 4139–4149. [CrossRef] [PubMed] 35. Pravdenkova, S.V.; Basnakian, A.G.; James, S.J.; Andersen, B.J. DNA fragmentation and nuclear endonuclease activity in rat brain after severe closed head injury. Brain Res. 1996, 729, 151–155. [CrossRef] 36. Schlade-Bartusiak, K.; Stembalska-Kozlowska, A.; Bernady, M.; Kudyba, M.; Sasiadek, M. Analysis of adaptive response to bleomycin and mitomycin C. Mutat. Res. 2002, 513, 75–81. [CrossRef] Int. J. Mol. Sci. 2020, 21, 8665 16 of 16

37. Apostolov, E.O.; Soultanova, I.; Savenka, A.; Bagandov, O.O.; Yin, X.; Stewart, A.G.; Walker, R.B.; Basnakian, A.G. Deoxyribonuclease I is essential for DNA fragmentation induced by gamma radiation in mice. Radiat. Res. 2009, 172, 481–492. [CrossRef] 38. Shiokawa, D.; Iwamatsu, A.; Tanuma, S. Purification, characterization, and amino acid sequencing of DNase gamma from rat spleen. Arch. Biochem. Biophys. 1997, 346, 15–20. [CrossRef] 39. Widlak, P.; Li, L.Y.; Wang, X.; Garrard, W.T. Action of recombinant human apoptotic endonuclease G on naked DNA and chromatin substrates: Cooperation with exonuclease and DNase I. J. Biol. Chem. 2001, 276, 48404–48409. [CrossRef] 40. Oliveri, M.; Daga, A.; Lunardi, C.; Navone, R.; Millo, R.; Puccetti, A. DNase I behaves as a transcription factor which modulates Fas expression in human cells. Eur. J. Immunol. 2004, 34, 273–279. [CrossRef] 41. Basnakian, A.G.; Singh, A.B.; Shah, S.V. Identification and expression of deoxyribonuclease (DNase) I alternative transcripts in the rat. Gene 2002, 289, 87–96. [CrossRef] 42. Takeshita, H.; Mogi, K.; Yasuda, T.; Nakajima, T.; Nakashima, Y.; Mori, S.; Hoshino, T.; Kishi, K. Mammalian deoxyribonucleases I are classified into three types: Pancreas, parotid, and pancreas-parotid (mixed), based on differences in their tissue concentrations. Biochem. Biophys. Res. Commun. 2000, 269, 481–484. [CrossRef] [PubMed] 43. Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227, 680–685. [CrossRef][PubMed] 44. Hofnagel, O.; Luechtenborg, B.; Stolle, K.; Lorkowski, S.; Eschert, H.; Plenz, G.; Robenek, H. Proinflammatory cytokines regulate LOX-1 expression in vascular smooth muscle cells. Arter. Thromb. Vasc. Biol. 2004, 24, 1789–1795. [CrossRef][PubMed] 45. Apostolov,E.O.; Ray,D.; Alobuia, W.M.; Mikhailova, M.V.; Wang, X.; Basnakian, A.G.; Shah, S.V.Endonuclease G mediates endothelial cell death induced by carbamylated LDL. Am. J. Physiol. Heart Circ. Physiol. 2011, 300, H1997–H2004. [CrossRef][PubMed]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).