The Occurrence of Single-Stranded DNA
`in the Serum of Patients with
`Systemic Lupus Erythematosus and Other Diseases
`
`DAVID KOFFLER, VINCENT AGNELLO, ROBERT WINCHESTER, and
`HENRY G. KUNKEL
`From the Department of Pathology of Mount Sinai School of Medicine, New
`York 10029, and The Rockefeller University, New York 10021
`
`Single-stranded DNA (SDNA) occurs
`A B S T R A C T
`in high incidence and in greatest concentration in the
`lupus erythematosus
`patients with systemic
`sera of
`(SLE), where levels as high as 250 /sg/ml were ob-
`served. SDNA appears to be an imunogen for anti-
`SDNA antibodies and forms complexes in vivo of both
`anti-SDNA-SDNA and anti-NDNA-SDNA types, which
`apparently play a role in the pathogenesis of the glomeru-
`lonephritis found in patients with SLE. SDNA is also
`found in high incidence but at lower levels in the sera
`of patients with rheumatoid arthritis. Lesser amounts
`of SDNA are found in several other diseases in which
`a low incidence of anti-SDNA antibodies is observed.
`
`INTRODUCTION
`that single-stranded
`Previous studies have indicated
`DNA (SDNA)1 is a potential immunogen for antibodies
`reactive with polynucleotides in various diseases (1, 2).
`The high incidence of anti-SDNA antibodies, especially
`in certain diseases with active tissue destruction (1),
`suggests that SDNA antigen may be frequently released
`into the circulation. Anti-SDNA antibody has been ob-
`served to occur with greatest frequency in patients with
`systemic lupus erythematosus (SLE). A question of ma-
`jor interest is whether these antibodies play a role in the
`pathogenesis of tissue lesions by combining with SDNA
`
`This work was supported by U. S. Public Health Service
`Grants AM 13721 and AM 04761.
`Received for publication 27 July 1972 and in revised form
`5 September 1972.
`1 Abbreviations used in this paper: anti-A, rabbit anti-
`sera prepared to adenosine; anti-T, rabbit antisera prepared
`to thymidine bovine serum albumin; NDNA, native DNA;
`SDNA, single-stranded DNA; SLE, systemic lupus erythe-
`matosus.
`
`and forming immune complexes. Studies of glomerular
`eluates obtained from SLE kidneys indicate that anti-
`SDNA antibodies are selectively concentrated *in the
`glomeruli of these kidneys (3), and immunofluorescence
`studies have demonstrated that SDNA antigen is de-
`posited in glomerular lesions of SLE kidneys in associ-
`ation with gamma globulin and complement deposits (4).
`The present investigation was undertaken to assess
`the occurrence of SDNA antigen in sera of patients with
`SLE and other diseases. The relationship between the
`appearance of SDNA antigen and antibody during the
`course of the disease was studied. Direct evidence for
`the presence of circulating immune complexes was also
`sought by assaying for the presence of both antigen and
`antibody in the same serum specimens. The evidence ob-
`tained from these studies indicates that serological as-
`pects of the SDNA-anti-SDNA system have similarities
`to those of the native DNA (NDNA)-anti-NDNA sys-
`tem. Anti-NDNA antibodies which also are found in
`high incidence in patients with clinically active SLE (1,
`2) have been demonstrated to alternate with DNA anti-
`gen in serum (5, 6), and most probably participate in
`the formation of antigen-antibody complexes. Consid-
`erable evidence has been accumulated indicating that im-
`mune complexes comprised of SDNA antigen and anti-
`body are also formed during the periods when patients
`with SLE have disease activity or progressive renal
`lesions.
`
`MATERIALS AND METHODS
`Human sera were obtained from patients with the following
`diseases: SLE (60), rheumatoid arthritis
`(54), chronic
`glomerulonephritis (40), leukemia (19), malignant tumors
`(20), random hospital diseases (60), and normal subjects
`(56). In addition, sera from serial studies of 18 patients
`
`198
`
`The Journal of Clinical Investigation
`
`Volume 52
`
`January 1973
`
`00001
`
`EX1042
`
`

`

`with SLE and from 10 patients with rheumatoid arthritis
`were assayed. The serial studies were comprised of 470 sera
`from SLE patients and 192 sera from rheumatoid arthritis
`patients.
`Polynucleotides employed for immunological tests. Na-
`tive calf thymus' DNA was obtained from Worthington
`Biochemical Corp., Freehold, N. J.; poly A, from Miles
`Laboratories, Inc., Kankakee, Ill.; and poly dA, dT, dG,
`and dC, from Biopolymers, Inc., Chagrin Falls, Ohio. De-
`natured DNA was prepared by heating native DNA at
`100'C for 10 min and transferring directly to an ice bath.
`Three sera were used for the routine assay of SDNA
`in serum: rabbit antisera prepared to adenosine (anti-A)
`and to thymidine bovine serum albumin (anti-T) (7), and
`an SLE serum (Esp). In addition, selected sera were tested
`for SDNA with a rabbit antiserum prepared to SDNA
`complexed to methylated bovine serum albumin (8). The
`latter sera showed a sensitivity similar to that of Esp
`serum for detecting antigen. They showed no reactivity
`with native DNA when tested by direct hemagglutination
`or agar gel diffusion. An SLE serum containing antibodies
`reactive with NDNA (Ann) was also used in the hemag-
`glutination inhibition test for determination of DNA.
`The specificities and relative sensitivities of the antisera
`were determined by a hemagglutination inhibition test using
`a constant end-point dilution of antiserum and serial dilu-
`tions of antigen (Table I). The rabbit anti-A and anti-T
`sera showed restricted specificity for homologous polynu-
`cleotides, whereas the SLE serum showed reactivity with
`most polynucleotides tested, indicating a considerably greater
`heterogeneity of antibody directed against determinants of
`SDNA. The rabbit anti-SDNA serum showed more re-
`stricted specificity, reacting with both dG and dC.
`Assay of sera for the presence of total DNA or SDNA
`was carried out by a hemagglutination inhibition test as
`previously described (2). A constant end-point dilution of
`an antiserum reactive with SDNA was used with serial
`dilutions of sera to be tested for SDNA. The maximal
`inhibition by anti-A, anti-T, or Esp serum of hemagglutina-
`tion of red blood cells coated with SDNA was compared
`with the inhibitory activity of a known concentration of
`SDNA, and the approximate concentration was calculated.
`(Esp) was
`sensitivity
`The antiserum with the greatest
`capable of detecting 0.4 ,g of SDNA in serum. The true
`incidence of NDNA antigen could not be directly ascer-
`tained by hemagglutination inhibition, inasmuch as insuf-
`ficient amounts of sera with unique specificity for NDNA
`were available. Human serum (Ann) was therefore utilized
`to determine "total" DNA, and sera reactive with Ann but
`inactive with anti-T, anti-A and Esp were considered to
`have NDNA.
`
`TABLE I
`Minimal Concentration of Polynucleotide Antigens Inhibiting
`End-Point Dilution of A nti-SDNA A ntisera
`
`Antiserum
`
`NDNA SDNA A dA
`
`dT dG dC
`
`SLE patient Esp
`Rabbit anti-A (no. 441)
`Rabbit anti-T (no. 440)
`Rabbit anti-SDNA (no. 285)
`
`-*
`-
`-
`-
`
`0.41
`31 - 5
`7.5
`15
`3.8 - - -
`15
`0.9
`1.9 - - 1.9 - -
`0.4 - - - 1.2
`0.6
`
`* No inhibition.
`$ Concentration of antigen in micrograms per milliliter.
`
`TABLE I I
`SDNA Antigen in Human Sera*
`
`Incidence of
`
`No. of
`sera
`
`SDNA
`antibody
`
`SDNA
`antigen
`
`SLE
`Rheumatoid arthritis
`Chronic glomerulo-
`nephritis
`Leukemia
`Malignant tumors
`Hospital diseases
`Normal human sera
`
`60
`54
`
`40
`19
`20
`99
`56
`
`% positive % positive:
`55
`52
`49
`24
`
`5
`5
`0
`5
`4
`
`18
`39
`15
`20
`4
`
`* Hemagglutination inhibition assay.
`Incidence in sera without detectable antibody.
`§ Mean concentration.
`
`SDNA
`
`ug/ml
`23.0§
`18.6
`
`5.9
`9.6
`7.5
`6.0
`5.6
`
`Agar gel diffusion was utilized as an ancillary test for
`detection of SDNA in serum (5). Rabbit anti-A and anti-T
`sera contained precipitating antibodies capable of detecting
`5-10 ,ug of SDNA, although clear precipitin reactions were
`demonstrable only at higher concentrations. This technique
`was therefore used to confirm the presence of SDNA
`demonstrated by hemagglutination inhibition in selected sera
`and to ascertain the antigenic identity of SDNA prepared
`in vitro and that detected in human serum.
`Deoxyribonuclease and ribonuclease
`Enzyme treatment.
`treatment of sera were performed as previously described
`(2). Diphenylamine determinations for DNA were per-
`formed as previously described (5).
`
`RESULTS
`Incidence of SDNA antigen in various diseases (Table
`Approximately one-half of SLE and rheumatoid
`II).
`arthritis sera negative for anti-SDNA antibody con-
`tained circulating SDNA antigen. A somewhat lower
`incidence of sera positive for SDNA antigen occurred in
`groups of diseases in which antibody was not detectable.
`SDNA appeared infrequently in normal sera. Sera from
`individual patients with SLE and rheumatoid arthritis
`manifested the highest levels of SDNA antigen. In con-
`trast, all other groups of sera contained SDNA anti-
`gen in the range of 5-10 ig/ml.
`DNA antigen in sera from serial studies of SLE pa-
`tients. Sera from a group of 18 patients with SLE were
`studied for periods of 6-51 mo for the presence of anti-
`SDNA antibody and SDNA antigen (Table III). 17 of
`18 patients manifested anti-SDNA antibodies during cir-
`cumscribed periods of time, with a widely varying fre-
`quency of occurrence in sera obtained at different stages
`of disease. SDNA antigen was demonstrated in sera ob-
`tained from 13 of 18 patients. In contrast to the moderate
`levels of SDNA present in randomly selected SLE sera,
`the serial studies revealed sera from certain patients with
`extremely high levels of SDNA in the range of 125-250
`,eg/ml. The mean serum concentration of SDNA antigen
`
`Single-Stranded DNA and SLE
`
`199
`
`00002
`
`

`

`TABLE I I I
`Incidence of SDNA Antigen and Antibody in Serial Studies of SLE Patients
`
`Patient
`
`Duration
`of study
`
`No. of
`sera tested
`
`% of sera
`positive for
`SDNA antigen
`
`Mean concn of
`SDNA antigen
`in sera
`
`% of sera
`positive for
`SDNA antibody
`
`Wol
`ILoo
`Esp
`Gam
`Har
`Vra
`Bro
`Hati
`Ros
`McK
`For
`EgI
`Ber
`Pea
`Rtis
`Sil
`Wvil
`Wei
`
`mo
`24
`6
`25
`9
`22
`36
`31
`51
`24
`16
`24
`31
`16
`17
`31
`13
`34
`10
`
`22
`11
`15
`21
`24
`45
`26
`55
`25
`19
`24
`34
`28
`24
`40
`15
`22
`20
`
`0
`0
`0
`0
`0
`13
`15
`15
`20
`21
`25
`28
`29
`33
`48
`60
`64
`90
`
`mg/ml
`0
`0
`0
`0
`0
`41.3
`17.1
`20.1
`19.9
`15.2
`19.1
`22.2
`18.2
`25.9
`65.4
`111.8
`125.4
`52.6
`
`23
`46
`54
`66
`42
`62
`39
`0
`32
`37
`54
`21
`32
`83
`32
`27
`27
`20
`
`in the serial studies was 53.0 isg/ml, more than twice
`that observed in the random SLE sera (Table IV).
`Five serial studies are cited below which illustrate the
`main types of data derived from these studies with re-
`spect to SDNA antigen. The majority of patients showed
`levels of SDNA antigen alternating with anti-SDNA
`antibody (Fig. 1). One patient (Sil) showed high levels
`of antigen ranging from 100 to 250 gg/ml. In certain
`patients, antigen in serum persisted for long periods of
`time and was associated with the presence of renal dis-
`ease. Patient Rus manifested a period of antigenemia
`associated with proteinuria followed by a quiescent pe-
`riod. A second episode of proteinuria was associated with
`high titers of anti-SDNA and anti-NDNA antibody
`and no detectable DNA antigen (Fig. 2). In contrast,
`certain patients demonstrated anti-SDNA antibody with
`only occasional sera containing detectable SDNA anti-
`gen. 7 of 23 sera obtained from a patient with active renal
`disease (Pea) contained antigen, whereas most sera
`
`TABLE IV
`Summary of Serial Studies
`
`demonstrated anti-SDNA antibody (Fig. 3). Five of the
`sera containing SDNA antigen were also shown to have
`antibody, indicating that imune complexes were present.
`Another patient (Vra) also demonstrated predominantly
`antibody activity in multiple sera (Fig. 4). Isolated sera
`containing antigen preceded or followed peaks of anti-
`body activity. It should also be noted that several peaks
`of NDNA antigen were observed in the absence of
`SDNA in sera that had been stored at 4VC for several
`years.
`No unique clinical or serological features were ap-
`parent in the patients in whom SDNA was not identified
`in any sera. Two of the five patients with no demon-
`
`260'
`220
`180
`SDNA
`antigen 140 -
`( ug/ml) 100I
`
`Sil
`
`I j..
`
`Number of patients
`Number of sera
`Number of sera containing SDNA
`Mean concentration of SDNA in
`positive sera (ug/nl)
`
`SLE
`
`18
`470
`118
`
`53.0
`
`Rheumatoid
`arthritis
`
`10
`192
`71
`
`13.6
`
`antibodies
`
`2E
`
`D
`
`F M
`J
`N
`1968
`1969
`Sil. Patient with SLE and progressive renal dis-
`FIGURE 1
`ease. SDNA antigen and antibody alternate in appearance,
`with extremely high levels of SDNA present during periods
`of antigenemia.
`
`200
`
`D. Koffler, V Agnello, R. Winchester, and H. G. Kunkel
`
`00003
`
`

`

`Rus
`
`l
`
`iI _
`
`-M]I
`
`..U| lJ
`
`,W 11
`"IT_'
`
`11Iu
`I IIh
`
`M
`
`J
`
`S
`
`N
`
`J S
`
`SDNA
`antigen
`(/Ig/ml)
`
`140 -
`looL
`
`20~ -Jf
`
`SDNA
`antibodies
`
`NDNA
`antibodies
`
`10
`6
`2
`
`12r
`8E
`41
`
`Ad
`
`Proteinuria
`(g /24 h
`
`~
`S
`
`N
`
`10
`6
`2y
`J M M
`MM J
`M
`J
`1970
`1968
`1969
`a period of 23A yr during which two major
`Rus. Patient with SLE followed for
`FIGURE 2
`episodes of glomerulonephritis occurred. The first exacerbation of renal disease was associated
`with a brief rise in titer of anti-NDNA antibodies and no demonstrable anti-SDNA anti-
`bodies. The second episode of renal disease was associated with high titers of both anti-SDNA
`and anti-NDNA antibodies. SDNA was detectable in sera during the first episode of renal
`disease, whereas none was observed during the second episode.
`
`strable SDNA (Loo and Esp) showed multiple sera
`containing NDNA. Patient Hau, who has been followed
`for a period of 51 mo with no demonstrable anti-SDNA
`antibody, had a relatively low level and incidence of
`SDNA antigen (15% of 55 sera), although this patient
`had severe renal disease and active SLE.
`Serial studies of sera from rheumatoid arthritis pa-
`tients also showed a widely varying frequency of oc-
`currence of SDNA antigen and antibody. 4 of the 10
`studies had only one serum positive for SDNA (Table
`V). The levels of antigen observed in these were con-
`siderably lower than in the SLE serial studies, and the
`mean concentration of SDNA was 13.6 iug/ml (Table
`IV).
`Agar gel diffusion.
`Sera containing demonstrable
`SDNA by hemagglutination inhibition were tested by
`agar gel diffusion. Sera with higher concentrations of
`SDNA gave clear precipitin reactions with rabbit anti-A
`or anti-T sera. The precipitin reactions between the
`rabbit antisera and the SLE sera showed lines
`of
`identity when in vitro prepared SDNA was placed in
`a well adjacent to the SLE sera. Sera with lower con-
`centrations of SDNA, i.e. less than 15 jg/ml, gave
`weak precipitin lines with rabbit antisera or were un-
`reactive in agar gel. An SLE serum (Rus) containing
`125 /ig/ml of SDNA as determined by hemagglutination
`5 a to precipitate with
`is shown in Fig.
`inhibition,
`rabbit anti-A serum. The rabbit antiserum demonstrates
`a precipitin reaction with in vitro prepared SDNA. An
`additional precipitin line between the SLE serum and
`SDNA is observed. This appears to be due to the
`presence of excess anti-NDNA antibodies because, as
`
`shown in Fig. 5 b, absorption with NDNA eliminates
`the precipitin line. These as well as other studies indi-
`cate that this SLE serum contains both SDNA with
`reactive sites for the rabbit antiserum and anti-NDNA
`antibodies reactive with NDNA and SDNA.
`DNase treatment of serum re-
`Enzyme treatments.
`sulted in a decrease of 50% or greater in hemagglutina-
`tion inhibitory activity of serum. 12 sera with SDNA
`inhibitory titers which averaged 5.3 (log base 2) were
`reduced to inhibitory titers of 2.5 by DNase treatment.
`A similar effect was observed when NDNA or SDNA
`was added to a normal serum with no previous inhibitory
`activity. Ribonuclease incubation did not affect the in-
`hibitory titers of any serum tested.
`Selected sera were as-
`Diphenylamine determination.
`sayed for the presence of DNA by this method. Five sera
`with concentrations ranging from 31 to 250 Ag/ml as de-
`termined by hemagglutination inhibition were found to
`
`SDNA.
`antigen
`(,ug/ml)
`
`60 E
`20L
`
`__ _I
`
`IL_
`
`Poo
`I I
`
`SDNA 6
`26
`antibodies
`
`fli I|.
`
`J M M J
`N
`1968 1969
`
`S
`
`N
`
`J M
`1970
`
`Pea. Patient with SLE and an episode of renal
`FIGURE 3
`disease who manifested clinical activity throughout most of
`the period of observation. Anti-SDNA antibodies were found
`in most sera, including five of seven sera in which SDNA
`antigen was demonstrable.
`
`Single-Stranded DNA and SLE
`
`201
`
`00004
`
`

`

`Clinical
`activity
`SDNA
`antigen
`( .tg/ml)
`
`80
`40
`
`OWL..
`
`. ..- 1.
`
`.
`
`1.
`
`Vra
`
`:oll I III .1
`
`A*.i..i...
`
`_1.
`
`I
`
`. .
`
`. ----A II
`
`8
`
`80
`40
`
`10
`
`62
`
`SDNA
`antibodies
`
`DNA
`antigen
`(Gg/ml
`DNA
`ontibodies
`
`I I 2 . . . --
`*111
`
`N
`
`J
`
`S
`
`N
`
`J
`
`i
`S N
`
`J
`
`M
`
`J
`
`.
`J M M
`M
`M
`1967
`1968
`1969
`1970
`Vra. Patient with SLE showing multiple episodes of clinical exacerbation. Anti-
`FIGURE 4
`SDNA antibody was present throughout most of the clinical course. SDNA antigen was
`demonstrable intermittently in a few sera, either before or after a peak of anti-SDNA
`antibody activity. The temporal relationship of SDNA antigen and antibody suggests that
`antigen-antibody complexes were formed during the first two clinical episodes. Note the
`occurrence of two periods when NDNA antigen alternates with the presence of anti-NDNA
`antibody in the absence of SDNA antigen.
`
`have concentrations ranging from 19 to 232 gg/ml by
`chemical assay.
`
`DISCUSSION
`Single-stranded DNA antigen has been found in the sera
`of patients with several diseases. The highest levels of
`antigen were found in patients with SLE, especially when
`sera were assayed throughout the course of the disease
`in individual patients. Sera obtained from patients with
`rheumatoid arthritis had moderately elevated levels of
`SDNA, whereas all other patients had low levels of this
`antigen. The incidence of antigen detection was highest
`in SLE and rheumatoid arthritis patients, although a
`
`significant number of sera from patients with other dis-
`eases contained small amounts of SDNA. These results
`support the hypothesis that SDNA is immunogenic in
`patients with SLE and rheumatoid arthritis, in which a
`high incidence of anti-SDNA antibodies has been ob-
`served (1, 2). In contrast, anti-SDNA antibodies occur
`in low incidence in those groups of patients in which
`lesser amounts of SDNA were detectable.
`The circulating SDNA antigen in SLE appears to
`have special significance in relation to the pathogenesis
`of the disease. The presence of antigen alternates with
`the appearance of anti-SDNA antibody in a fashion simi-
`lar to that demonstrated for the NDNA-anti-NDNA
`
`TABLE V
`Incidence of SDNA Antigen and Antibody in Serial Studies of
`Rheumatoid Arthritis Patients
`
`Patient
`
`Duration
`of study
`
`No. of
`sera tested
`
`% of sera
`positive for
`SDNA antigen
`
`Mean concn of
`SDNA antigen
`in sera
`
`% of sera
`positive for
`SDNA antibody
`
`Joh
`Los
`Qui
`Kov
`Rom
`Hil
`Mar
`Fal
`Coo
`Gar
`
`mo
`20
`37
`46
`12
`8
`26
`18
`58
`5
`30
`
`3 1
`37
`16
`9
`17
`29
`12
`17
`9
`15
`
`3
`3
`6
`11
`47
`48
`75
`76
`78
`100
`
`Ag/mt
`15.0
`31.0
`3.8
`15.0
`14.3
`16.5
`12.5
`15.1
`7.5
`11.5
`
`35
`86
`12
`45
`29
`0
`17
`12
`22
`0
`
`202
`
`D. Koffler, V. Agnello, R. Winchester, and H. G. Kunkel
`
`00005
`
`

`

`system (5, 6), suggesting that SDNA-anti-SDNA im-
`mune complexes are formed in serum. Additional evi-
`dence for in vivo antigen-antibody complex formation
`was obtained from the simultaneous demonstration of
`antigen and antibody in several sera as illustrated in
`the study of patient Pea, and from serial studies which
`indicate that SDNA antigen excess may immediately
`precede or follow peaks of anti-SDNA antibody activity
`as observed in patients Sil and Vra.
`The association of rises in the titers of anti-SDNA
`antibody with serum complement depression and clinical
`activity is less frequent than that observed for anti-NDNA
`antibodies. These antibodies, however, frequently rise in
`parallel with anti-NDNA during clinically active phases
`of the disease. SDNA antigen was usually found to
`appear in groups of sera obtained from circumscribed pe-
`riods during the course of the disease. In certain patients,
`a more prolonged period of persistence of SDNA antigen
`was observed extending over periods of 4-6 mo. In sev-
`eral patients, e.g., Rus, clinical exacerbations were as-
`sociated with the appearance of SDNA antigen, and
`patients with
`disease occurred in
`progressive
`renal
`persistently large amounts of circulating SDNA antigen.
`that
`indicate
`eluates
`of glomerular
`studies
`Recent
`
`FIGURE 5 a Agar gel precipitin reaction between (a) rab-
`(b) an SLE serum (Rus), and (c)
`bit anti-A serum,
`SDNA antigen at 250 9g/ml prepared in vitro. The highly
`specific anti-adenosine antiserum shows a precipitin reaction
`with the SDNA preparation and the SLE serum. There
`is also a precipitin band between the SDNA preparation
`and the SLE serum. This serum shows the simultaneous
`presence of SDNA and antibodies to NDNA.
`
`FIGURE 5 b Agar gel precipitin reaction with same reac-
`tants as in Fig. 5 a. After addition of NDNA antigen to the
`line between this serum and
`SLE serum, the precipitin
`SDNA is no longer demonstrable.
`
`SDNA-anti-SDNA immune complexes are deposited in
`renal tissues, and are the subject of a separate report.
`Therefore, evidence from both serological and tissue
`studies suggests that these complexes play a pathogenic
`role in the renal disease associated with SLE.
`Agar gel diffusion studies indicate that anti-NDNA-
`SDNA complexes are also present in the sera of SLE
`in which SDNA
`(Rus),
`one serum
`In
`patients.
`antigen was found, precipitating anti-NDNA antibodies
`were noted. Previous studies have shown that these anti-
`bodies were more reactive with SDNA, indicating that
`native sites on the SDNA molecules were saturated
`with antibody. It appears that three types of complexes
`may be formed in SLE sera: SDNA-anti-NDNA,
`SDNA-anti-SDNA, and NDNA-anti-NDNA. The rel-
`ative cytotoxicity of each of these types of complexes
`has not been ascertained, although all three types appear
`to be formed in SLE sera.
`The demonstration of SDNA in human serum raises
`several questions concerning the nature and source of the
`material. First, the demonstrated antigenic determinants
`appear to be present on predominantly single-stranded
`DNA and not present on cross-reacting single-stranded
`or double-stranded RNA. The susceptibility of the ma-
`terial to DNase treatment and the failure of RNase to
`decrease the inhibitory titer, the presence of thymidine
`identified with specific rabbit anti-sera, and the lines of
`identity with SDNA in agar gel diffusion indicate that
`SDNA is present in serum. Chemical assays with di-
`phenylamine also are indicative of the presence of DNA,
`although these do not differentiate native from single-
`stranded DNA. The failure of DNase to completely de-
`stroy the inhibitory activity of the material was paralleled
`by the inability of the enzyme to completely remove the
`inhibitory activities of both native and single-stranded
`DNA added in vitro to normal sera. The breakdown
`products of DNA after DNase appear to retain some
`reactivity with antibody, although to a significantly lesser
`degree.
`The presence of large quantities of SDNA in the sera
`of several patients suggests that the antigen was mainly
`of host origin. Studies are in progress to determine if
`the DNA found in these sera is more closely homologous
`to mammalian rather than viral DNA. The present stud-
`ies indicate that the NDNA is not denatured in vitro.
`Several sera refrigerated at 4VC for periods of months
`retained the reactivity of NDNA with no evidence of
`SDNA formation. Previous investigations of NDNA in
`sera indicated that this material was relatively stable in
`serum and that the small amount of serum deoxyribonu-
`clease was ineffective in causing the breakdown of DNA
`in serum. These studies, however, do not exclude the
`possibility that at least part of the SDNA is of exogenous
`origin, nor do they indicate whether SDNA is released
`
`Single-Stranded DNA and SLE
`
`203
`
`00006
`
`

`

`directly from areas of tissue injury or formed in vivo
`from NDNA. Further immunofluorescence studies will
`be required to determine if DNA released at sites of
`tissue injury, such as skin (9), is composed of NDNA
`or SDNA. The influence of steroid therapy on the re-
`lease of NDNA into the circulation has been cited (10).
`In the present study, no direct correlation between ster-
`oid treatment and the appearance of large amounts of
`SDNA in the circulation was observed.
`DNA-"like" material has previously been found in
`cryoglobulins isolated from the sera of patients with
`mixed cryoglobulinemia (11). This material was not
`susceptible to DNase treatment at neutral pH, although
`precipitin reactivity was eliminated by treatment with
`DNase at acid pH. Reactions of partial identity were
`observed with SDNA. Cryoproteins obtained from rheu-
`matoid as well as nonrheumatoid synovial fluid have been
`shown to contain DNA (11, 12). No association between
`the presence of DNA and anti-NDNA or anti-SDNA
`antibody in the cryoproteins has been demonstrated. The
`pathogenetic significance of DNA in cryoproteins re-
`mains to be clarified by further study. In the present
`investigation, SDNA was also found in sera of patients
`with rheumatoid arthritis. Although these patients, simi-
`lar to patients with SLE, have a high incidence of anti-
`SDNA antibodies, SDNA was present in lesser amounts
`in sera of serial studies of rheumatoid arthritis. In cer-
`tain SLE patients, levels of 125-250 ug/ml of SDNA
`were observed, whereas the highest levels of SDNA in
`rheumatoid arthritis sera were 30-62 Ag/ml. The all-
`ternating appearance of antigen and antibody, which was
`frequently evident in SLE patients, was infrequent in
`of the
`patients with rheumatoid arthritis. The role
`SDNA-anti-SDNA system in rheumatoid joint fluids re-
`mains to be explored, but humoral SDNA-anti-SDNA
`complexes do not appear to exert any pathogenic effect
`on renal glomeruli or blood vessels in patients with rheu-
`matoid arthritis.
`The results of this study, therefore, indicate that cir-
`culating SDNA occurs at higher levels in patients with
`SLE than with other diseases, in contradistinction to
`to NDNA, which has been found at high levels in pa-
`tients with other diseases (13). The accumulated evi-
`dence indicates that it participates in the formation of
`immune complexes and may be an immunogen for SDNA
`antibody. Its role in rheumatoid arthritis is less clear,
`although it is a likely immunogen in these patients as
`
`well, whereas the presence of small amounts of SDNA
`antigen in other diseases such as leukemia is probably
`a reflection of nonspecific tissue breakdown.
`
`ACKNOWLEDGMENT
`The authors are greatly indebted to Miss Lydia Legrand
`for her excellent technical assistance.
`
`REFERENCES
`1. Koffier, D., R. I. Carr, V. Agnello, T. Fiezi, and H. G.
`Kunkel. 1969. Antibodies to polynucleotides: distribution
`in human serums. Science (Wash. D. C.). 166: 1648.
`D., R. Carr, V. Agnello, R. Thoburn, and
`2. Koffler,
`H. G. Kunkel. 1971. Antibodies to polynucleotides in
`human sera: antigenic specificity and relation to disease.
`J. Exp. Med. 134: 294.
`3. Koffler, D., V. Agnello, R. Thoburn, and H. G. Kunkel.
`1971. Systemic lupus erythematosus: prototype of im-
`mune complex nephritis in man. J. Exp. Med. 134(3,
`Pt. 2): 169s.
`4. Andres, G. A., L. Accinni, S. M. Beiser, C. L. Christian,
`G. A. Cinotti, B. F. Erlanger, K. C. Hsu, and B. C.
`1970.
`fluorescein-labeled
`Localization of
`anti-
`Seegal.
`nucleoside antibodies in glomeruli of patients with active
`systemic lupus erythematosus nephritis. J. Clin. Invest.
`49: 2106.
`5. Tan, E. M., P. H. Schur, R. I. Carr, and H. G. Kunkel.
`1966. Deoxyribonucleic acid (DNA) and antibodies to
`DNA in the serum of patients with systemic lupus
`erythematosus. J. Clin. Invest. 45: 1732.
`6. Schur, P. H., and J. Sandson. 1968. Immunological
`factors and clinical activity in systemic lupus erythema-
`tosus. N. Engl. J. Med. 278: 533.
`7. Erlanger, B. F., and S. M. Beiser.
`1964. Antibodies
`specific for ribonucleosides and ribonucleotides and their
`reaction with DNA. Proc. Nat!. Acad. Sci. U. S. A.
`52: 68.
`8. Plescia, 0. J., W. Braun, and N. C. Palczuk. 1964.
`Production of antibodies to denatured deoxyribonucleic
`acid (DNA). Proc. Natl. Acad. Sci. U. S. A. 52: 279.
`9. Tan, E. M., and H. G. Kunkel. 1966. An immunofluores-
`cent study of the skin lesions in systemic lupus erythe-
`matosus. Arthritis Rheurm. 9: 37.
`10. Hughes, G. R. V., S. A. Cohen, R. W. Lightfoot, Jr.,
`J. I. Meltzer, and C. L. Christian 1971. The release of
`DNA into serum and synovial fluid. Arthritis Rheum.
`14: 259.
`11. Barnett, E. V., R. Bluestone, A. Cracchiolo, III, L. S.
`Goldberg, G. L. Kantor, and R. M. McIntosh. 1970.
`Cryoglobulinemia and disease. Ann. Intern. Med. 73: 95.
`12. Marcus, R. L., and A. S. Townes. 1971. The occurrence
`of cryoproteins in synovial fluid; the association of a
`complement-fixing activity in rheumatoid synovial fluid
`with cold-precipitable protein. J. Clin. Invest. 50: 282.
`13. Carr, R. I. 1969. DNA antibodies and DNA in serum.
`Ph.D. Thesis. The Rockefeller University, New York.
`
`204
`
`D. Koffler, V. Agnello, R. Winchester, and H. G. Kunkel
`
`00007
`
`

`

`THE,JOURNAL OE
`ULI NIC Al
`‘ ■ i
`#
`TN VRSTI
`
`.
`
`■ - ....'J,,•>((■.ikfliMiw
`
`1
`
`IO N
`
`■»<
`
`f PUBLISHED BY THE AMERICAN SOCIETY FOR CLINICAL
`
`INVESTIGATION, INC.
`
`JANBS W3
`rv CJ '^ALIFOOIA
`
`JANUARY 1973 VOLUME 52 NUMBER 1
`
`*
`
`■'1
`
`JCIHAO 52(1) 1-222 (1973)
`
`00008
`
`