IUBio

Humanized, Reshaped, Humouse antibodies?

Mike Clark mrc7 at cam.ac.uk
Thu May 6 03:33:34 EST 1999


In article <37308C09.1C08 at mail.tju.edu>, mark
<URL:mailto:mark.haynes at mail.tju.edu> wrote:
> Mike Clark wrote:
> > 
> > I am about to write a review on the topic of antibody humanization and
> > would be interested in any views, comments, on my thoughts below.
[snipped]
> 
> 
> Great>  I have wondered about that too.

Mark thanks for your comments but I will try to expand a little on my first
posting whilst answering your reply.

> What is the status of human  antibodies as opposed to humanized ones?

The first problem I have is the definition of a human antibody versus a
humanized antibody. If I isolated a rodent antibody and found that the CDRs
were identical to an isolated human antibody but the frameworks were
different would a reshaped or CDR grafted version of the rodent antibody
which used the same frameworks as the other be human? I would claim I
think quite reasonably that it is because the sequences of the reshaped or
humanized antibody is identical to the 'isolated human' antibody. Clearly
of course I could also quite reasonably claim that it is a humanized
antibody because that is how I made it. So in some circumstances humanised
antibodies can also be human antibodies!

Alternatively take a different example. I isolate a human antibody from one
individual human  and put it into another human who lacks the V-gene which
encodes the antibody (such V-genes can be readily identified from the
Tomlinson 'V-Base' directory). The antibody is both human but it is also
'foreign' to the recipient. Now is it more or less 'foreign' than a
humanised rodent antibody which differed by the same number of amino acids
from the closest V-gene in the recipient?

And another example. Take an antibody made by a human B-cell which has
undergone somatic mutation and put it into an identical twin in which the
same somatic mutation may not have occurred. The antibody can again be
defined as both 'human' and yet is still 'foreign' to the recipient at the
same time. ie 'human' does not mean the same as 'self'.

Can you see the points I am trying to make? The definitions of human,
rodent, humanized, reshaped, foreign etc (all commonly used terms in the
field of antibody engineering) merely refer to the process and the context
by which the antibodies are produced and used. 

However what we really want to know is are they immunogenic? The point I am
trying to get at is that this is not necessarily dependent upon how the
antibody is made.

> Are phage libraires available  that are totally human sequences?

Yes there are. 

But if you then vary the CDR sequences in the library to mimic somatic
mutation are they still 'totally human sequence'?

> What results are known from the studies with the anti-TNF antibodies?

They are immunogenic in at least some of the recipients.

To expand further the rat IgG2b CD52 antibody CAMPATH-1G is very
immunogenic in most patients.. The reshaped (CDR-grafted) human IgG1
antibody is much less immunogenic (Ah success, humanisation works!)

> What about all the IVIG that has been given.

Many would claim that at least some of the actions of IVIG can be
attributed to idiotype antidiotype interactions!

> Are there any responses to  these therapies? markH
> 

Again I make the point that some chimaeric antibodies are not very
immunogenic in humans whereas others are. Is this merely a reflection of
how 'human' the sequences are or is another possibility that the immune
system of the recipient sees some antibodies as 'less dangerous' than
others? Again I make the point I made in my last posting that this concept
of purely 'self versus non-self' discrimination as a basis for antibody
humanisation may be fundamentally flawed if you don't also take into
account the concept of 'danger versus non-danger'.

If I wanted to raise a rodent anti-idiotype antibody to a rodent monoclonal
antibody there are many immunisation protocols which I could use which are
likely to work [eg use an adjuvant or couple the antibody to a protein such
as KLH etc etc]. (The last method is how we made an anti-idiotype to our
rat monoclonal antibody CAMPATH-1G). The trick is to make the antibody look
'dangerous'.



Mike Clark,                        <URL:http://www.path.cam.ac.uk/~mrc7/>
-- 
 o/ \\    //            ||  ,_ o   M.R. Clark, PhD. Division of Immunology
<\__,\\  //   __o       || /  /\,  Cambridge University, Dept. Pathology
 ">    ||   _`\<,_    //  \\ \> |  Tennis Court Rd., Cambridge CB2 1QP
  `    ||  (_)/ (_)  //    \\ \_   Tel.+44 1223 333705  Fax.+44 1223 333875







More information about the Immuno mailing list

Send comments to us at biosci-help [At] net.bio.net