|
|
 Previous Article | Table of Contents | Next Article 
Blood, 15 April 2002, Vol. 99, No. 8, pp. 3072-3074
CORRESPONDENCE
To the editor:
Multiple myeloma: illegitimate switch recombinations and their
relation to chromosomal translocations
We read with interest the article by Ho et
al1 on illegitimate switch recombinations in multiple
myeloma (MM). In this study a Southern blot method, established by
Bergsagel et al,2 was applied to detect rearrangements in
immunoglobulin heavy chain (IgH) switch regions. Five pairs of probes
that hybridize upstream or downstream of each switch region were used
on restriction digests of enzymes with digestion sites outside the pair
of switch probes. Based on the hybridization characteristics of these
probes legitimate and illegitimate switch recombination events can be
distinguished; the latter may be indicative of chromosomal
translocations. This assay was utilized by the authors to establish the
frequency of illegitimate rearrangements as an indicator of chromosomal
translocations in MM. We would like to make the following
comments: (1) As initially described the appropriate restriction enzyme digests
were used for the investigation of the individual switch regions (ie,
HindIII for switch (S) µ and S , SphI for
Sµ, S , and S , and BglII for Sµ, S µ, and
S ).2 In addition, other combinations of these
enzymes and probes (BglII/S, BglII/S, and
SphI/S, only the hybridization patterns of the 3'S and
3'S probe are shown) were applied in this study to confirm the
(il)legitimate nature of single rearranging bands in more than 1 enzyme. Not all of these combinations may be suitable for the
investigation of individual switch regions in terms of large-scale
screening: the pairs of switch probes were initially designed to
recognize the same germline fragment(s); otherwise, in the case of
restriction sites between both probes, the conclusiveness of the assay
is limited.2 With respect to the S regions on
BglII digests, our own results showed comigrating as well as
noncomigrating bands detected by the 5' and 3'S probe, respectively,
suggesting BglII digestion sites in at least 1 S
region.3 Thus, it should be shown whether or not the
corresponding 3' and 5' probes recognize the same germline fragments in
the above-mentioned combinations; if not, some legitimate
rearrangements might be misinterpreted as illegitimate due to the
separation of the corresponding 3' and 5' regions. (2) Figure 2A of Ho et al1 shows a 4.9 kilobase pair (kbp) SphI fragment detected by both
the 3' and 5'S probe that is interpreted as a translocation upstream
of 5'S or downstream of 3'S. As confirmation of this
rearrangement, a 5.2 kbp BglII fragment is presented that
hybridizes with the 3'S probe but not the 5'S probe. If
BglII encompasses both 5' and 3'S regions (a prerequisite
for an enzyme being used in this assay), the hybridization characteristics of the SphI and BglII fragments
are mutually exclusive because the BglII rearrangement is
detected by only 1 of the S probes. Therefore, either the fragments
are independent of each other or there is a BglII site
within this S region, meaning that both rearrangements may be due to
the same recombination/translocation outside of this S region. In
any case, it should not be concluded that a translocation in switch has been confirmed by 2 restriction enzymes. (3) Figure 2A of Ho et al1 shows 9.4 kbp
and 2.0 kbp BglII fragments detected by the 5'µ probe.
The latter was also detected by the 5'Sµ probe. In addition, 10 and
6.5 kbp HindIII bands hybridizing with the 5µ probe are
presented as confirmation of the former illegitimate rearrangements.
The 2.0 kbp BglII rearrangement detected by the 5'µ and
5'Sµ probe strongly suggests a recombination event downstream of the
5'Sµ region. The combination HindIII/5'µ, however, covers the region upstream of 5'Sµ including the joining region (Jh).
Thus the 2.0 kbp BglII fragment must be independent of the HindIII fragments mentioned. Alternatively, these
HindIII rearrangements could be due to VDJ
recombination events, which should have been tested with a Jh probe
(Bergsagel et al2[Fig2] for restriction map and positions
of probes). (4) In Figure 2A,C Ho et al1 show 12 kbp and 3.5 kbp
HindIII fragments detected by the 3'S probe. These
rearrangements were designated illegitimate recombination events as
they were not detected by probes for upstream or downstream acceptor
sites. However it is not mentioned whether or not these fragments
hybridized with the 5'S probe. This information is crucial to
exclude known HindIII restriction fragment length
polymorphisms (RFLPs) of the switch region.2,4 In summary we wish to point out that the true nature of
IgH-associated rearrangements remains obscure until the fragments have been cloned. Illegitimate rearrangements are indicative of, but do
not necessarily prove, chromosomal translocations; they may also be due
to polymorphisms and intrachromosomal rearrangements.2 Thus in terms of large-scale screening the true frequency of
translocations may be overestimated. In our experience the choice of
suitable alternative restriction enzymes for confirmation of single
IgH-associated (il)legitimate rearrangements has been complicated by
the high frequency of RFLPs in certain enzyme digests, digestion sites within switch regions, as well as unavailability of reliable DNA sequence data for restriction mapping. If, in the context of the above-mentioned assay, alternative restriction enzyme combinations are
used for the investigation of switch regions, additional
information should be given as to whether or not the corresponding 3'
and 5' switch probes recognize the same germline fragment. In any case
the results should be interpreted with care.
Helmut H. Schmidt
Correspondence: Helmut H. Schmidt, Division of Internal
Medicine, University of Graz, Department of Hematology,
Auenbruggerplatz 38, Graz, Austria A-8036
Acknowledgments
Supported in part by the Austrian Science Fund grant no. J1692.
References
1.
Ho PJ, Brown RD, Pelka GJ, Basten A, Gibson J, Joshua DE.
Illegitimate switch recombinations are present in approximately half of primary myeloma tumors, but do not relate to known prognostic indicators or survival.
Blood.
2001;97:490-495[Abstract/Free Full Text].
2.
Bergsagel PL, Chesi M, Nardini E, Brents LA, Kirby SL, Kuehl WM.
Promiscuous translocations into immunoglobulin heavy chain switch regions in multiple myeloma.
Proc Natl Acad Sci USA.
1996;93:13931-13936[Abstract/Free Full Text].
3.
Chen W, Palanisamy N, Schmidt H, et al.
Deregulation of FCG2RB expression by 1q21 rearrangements in follicular lymphomas.
Oncogene.
2001;20:7686-7693[CrossRef][Medline]
[Order article via Infotrieve].
4.
Bottaro A, Gallina R, DeMarchi M, Carbonara AO.
Genetic analysis of new restriction fragment length polymorphisms (RFLP) in the human IgH constant gene locus.
Eur J Immunol.
1989;19:2151-2157[Medline]
[Order article via Infotrieve].
Response:
Illegitimate switch recombinations in primary myeloma tumor
We thank Dr Schmidt for his comments on our study of
illegitimate IgH switch recombinations in primary myeloma
tumor.1 The first concern the letter raises relates to our
use of combinations of switch probes and restriction enzymes
"alternative" to those originally designed to hybridize the 5' and
3' ends of the switch regions, with restriction sites located outside
the probes (HindIII - 5'Sµ, 3'Sµ, 5'S, 3'S;
SphI - 5'Sµ, 3'Sµ, 5'S, 3'S, 5'S, 3'S; and
BglII - 5'Sµ, 3'Sµ, 5'µ,
3').2 First, we wish to emphasize (as
stated in "Materials and methods"1) that the strategy by which we screened patients for illegitimate recombinations, and by which all recombinant bands designated as illegitimate were
initially detected, utilized the above enzyme/probe combinations. We
agree that cloning would be most accurate in determining the nature of
these fragments. However, whereas myeloma cell lines (from which the
majority of translocation breakpoints have been cloned) and other
lymphoid tumor biopsies may provide abundant DNA for cloning, cell
numbers in marrow biopsies are often very limited. We therefore adopted
the approach of determining whether a recombinant fragment was
legitimate or illegitimate by "matching" with other switch probes.
As stated in "Materials and methods,"1 Southern hybridizations were performed with alternative enzyme/probe combinations, which had previously revealed recombinant fragments during screening. The basis of this approach was that if a fragment resulted from a legitimate recombination between 2 switch regions and
was detected by 1 of the 2 switch probes, then a matching fragment
should be detected by the other probe, provided that there are no
restriction sites internal to the probes. From the sequence data
available, we confirmed that there are no such internal restriction
sites for HindIII - 5'S, 3'S;
SphI-3'S; and BglII - 5'S, 3'S that
were required for our analysis. For the remaining combination
BglII-3'S, there are no BglII sites in S2,
3, and 4, but 1 site is present in S1. In our cohort,
BglII-3'S hybridization was performed to verify the
nature of nongermline BglII bands detected by Sµ probes in
samples also demonstrating HindIII-3'S nongermline
fragments. The difficulty in interpreting these BglII digests occurs when a legitimate Sµ to S1 switch may be considered illegitimate due to BglII digestion in S1. For all cases
in which BglII-3'S hybridization was done to "match"
with BglII-Sµ fragments, we emphasize the following: (1)
HindIII-5'Sµ, 3'Sµ, and 3'S were also performed; (2)
in all of these cases but 1 (Patient 1 discussed below), we interpreted
the results of both enzymes such that no BglII-Sµ fragment
was deemed illegitimate without confirmation with HindIII;
(3) and BglII-3'S remains informative for the other 3 S regions. As we discussed in the paper, we are fully aware that
some unmatched fragments could have resulted from polymorphisms, internal rearrangements, and deletions. We would like to point out that
in each hybridization we can detect the germline band(s) of expected
size, either from the nontumor cells in the marrow sample or the
nontranslocated allele, serving as an additional internal control.
Importantly, alternative enzyme/probe hybridizations were performed in
our study according to recombinant bands already demonstrated with the
"original" combinations and were used primarily as confirmation of
the screening Southern blots. These alternative combinations were
useful in eliminating undetected legitimate (especially downstream and
inversion) switches, and possible artifacts that can be introduced by
restriction fragment length polymorphisms (RFLPs) even in the original
screening hybridizations. The second criticism of Dr Schmidt relates to our analysis of 1 of 4 examples in Figure 2A (Patient 1).1 We agree that if a single translocation accounts for the 4.9 kilobase (kb)
SphI fragments detected by 5'S and 3'S, and
the 5.2 kb BglII-3'S fragment, then a BglII
recombinant band should also be detected by 5'S. However
BglII-5'S is clearly germline. From the sequence database, we verified for both S1 and 2 that a BglII site
is present 1.3 kb downstream of the 3' S probe (292 base
pair upstream of the 3' SphI site), and there are
no other BglII sites in S. From our germline
hybridization, upstream BglII sites appear to be located at
least 12 kb from the 3' site. Hence the 5.2 kb BglII fragment detected by 3'S is most likely to be caused by an
extraneous BglII site on a translocated fragment at the 5'
end, replacing the 5'S probe sequence. In the absence of other
BglII sites in S, we agree with Dr Schmidt that based on
our findings, the SphI-5'S and 3'S recombinant
fragments, although of the same size, must be independent. However this
does not reduce the possibility as we had suggested that the 5.2 kb
BglII and the 4.9 kb SphI 3'S fragments most
likely represent the same illegitimate recombination, although this
could only be formally proved by cloning. We have repeatedly
demonstrated the SphI-5'S fragment and the absence of
matching nongermline SphI fragments hybridized by 3'S and 5'Sµ, excluding downstream or inversion switches. We accept the criticism that the HindIII 5'µ recombinant fragments
cannot be used as confirmation of the 9.4 kb BglII fragment,
as 5'µ is located upstream of the HindIII site, a fact
that we had specified in the legend1(Fig2Aiii) but
overlooked in the analysis. Nevertheless Dr Schmidt agrees with us that
the 2 kb BglII fragments hybridized by both 5'µ and
5'Sµ strongly suggest a recombination event downstream of 5'Sµ.
From this analysis we believe that our data remain fully consistent
with the presence of illegitimate recombinations in this patient, and
our conclusions on the relationship with disease behavior are
unchanged. Regarding HindIII RFLPs in S, we wish to point
out that HindIII digests from all patients were hybridized by both 5'S and 3'S, which would have shown if any of the
extraneous bands were due to RFLPs. For the 12 kb and 3.5 kb
HindIII-3'S bands1(Fig2A,C) we can confirm
that no such bands were detected by 5'S. As a result, while cloning provides the ultimate proof of the nature of
a recombinant fragment in the IgH genes, we believe that our
investigation of illegitimate switch recombinations by the adaptation
of an established Southern Blot assay was justified, given the paucity
of available tumor material from human myeloma bone marrow. Our
exhaustive analysis using the original screening blots for detection of
possible illegitimate switches and their verification by probes and
enzymes that have previously revealed recombinant bands have provided
us with useful and reliable information. Since our paper was published
more than a year ago, substantial refinements have been made to
molecular cytogenetics,3,4 which are likely to be quicker,
less labor intensive and more accurate in demonstrating chromosome 14q
translocations, and would be superior to Southern hybridization for
large-scale patient screening. Finally, we would like to re-emphasize
that our conclusions regarding illegitimate switch recombinations were
made after careful investigation of each nongermline band. We fully
agree that detailed analysis is required in the interpretation of the
Southern Blot assay and thank Dr Schmidt for his note of caution.
P. Joy Ho, Ross D. Brown, Antony Basten, John Gibson, and Douglas E. Joshua
Correspondence: P. Joy Ho, Institute of Hematology, Royal Prince
Alfred Hospital, Missenden Rd, Camperdown, NSW, Australia 2050
References
1.
Ho PJ, Brown RD, Pelka GJ, Basten A, Gibson J, Joshua DE.
Illegitimate switch recombinations are present in approximately half of primary myeloma tumors, but do not relate to known prognostic indicators or survival.
Blood.
2001;97:490-495.
2.
Bergsagel PL, Chesi M, Nardini E, Brents LA, Kirby SL, Kuehl WM.
Promiscuous translocations into immunoglobulin heavy chain switch regions in multiple myeloma.
Proc Natl Acad Sci USA.
1996;93:13931-13936.
3.
Avet-Loiseau H, Daviet A, Brigaudeau C, et al.
Cytogenetic, interphase, and multicolor fluorescence in situ hybridization analyses in primary plasma cell leukemia: a study of 40 patients at diagnosis, on behalf of the Intergroupe Francophone du Myelome and the Groupe Francais de Cytogenetique Hematologique.
Blood.
2001;97:822-825[Abstract/Free Full Text].
4.
Sawyer JR, Lukacs JL, Thomas EL, et al.
Multicolour spectral karyotyping identifies new translocations and a recurring pathway for chromosome loss in multiple myeloma.
Brit J Haematol.
2001;112:167-174[CrossRef][Medline]
[Order article via Infotrieve]
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
Related Article in Blood Online:
-
Illegitimate switch recombinations are present in approximately half of primary myeloma tumors, but do not relate to known prognostic indicators or survival
- P. Joy Ho, Ross D. Brown, Gregory J. Pelka, Antony Basten, John Gibson, and Douglas E. Joshua
Blood 2001 97: 490-495.
[Abstract]
[Full Text]
[PDF]
|
|
