Thread by @DavidSteensma

#HematologyTweetstory 21: the “CD” nomenclature. Musical CDs may be obsolete (though I still have a lot of them from college days; Paul Simon’s “Graceland” wins for most listens). Communicating clearly about blood cell surface markers? Here to stay.🙂 @ASH_Hematology #hemepath/1

In the 1970s, German Georges Köhler & Argentinian-born César Milstein at @MRC_LMB fused myeloma cell lines w/ B cells to create immortal “hybridomas” that could produce specific monoclonal antibodies. Development of MoAbs led to a wave of discovery of cell-identifying markers./2

In the late 1970s & early 1980s, labs all across the world used this new #monoclonalantibody technology to move beyond morphology in cell identification. The same cell surface molecule identified by multiple laboratories was given different names. It got hard to read papers!/3

For example, in 1980 Lee Nadler & the late Stu Schlossman @DanaFarber used B-lymphocyte surface (BLys) antibody to identify antigen “B1” - and also found B2, B3, B4 and B5. But others called B1 Bp35, LEU-16, Ly-44, MS4A1, S7. (B1 is now CD20; B2 CD21, B3 CD22, B4 CD19, B5 CD25.)

This story was repeated over & over. The marker now called CD5 was described as T1, Tp67, Leu-1, & Ly-1. CD10 was CALLA (i.e., "Common ALL Antigen", since many #ALL cases express it), NEP, gp100, EC 3.4.24.11, etc. You can see the problem when trying to do a literature search!/5

In 1984, Alain Bernard @GustaveRoussy & Laurence Boumsell @INSERM in Paris wrote, “Over a small number of years many monoclonal antibodies directed against leucocytes have been produced. Their multiplicity & the various nomenclatures utilized may be disconcerting.” True, that./6

So in order to finally try to bring some order out of the chaos, a series of “Human Leucocyte Differentiation Antigens (#HLDA) Workshops” were created, where groups could compare their antibodies and see if they bound the same antigen. The first one was held in Paris in 1982./7

Ten #HLDA workshops have been held. Two of them were here in Boston: the 2nd in 1984 & 5th in 1993. Most recent was in 2014 in Wollongong, Australia, which brought the number of CD assignments up to CD371, including >400 molecules. Here are summaries of the first 2 workshops./8

Why were the antigens called “Clusters of Differentiation” (CD)? Data were analyzed statistically via
"cluster analysis"; this method tried to identify clusters/groups of antibodies with similar patterns of leukocyte binding at various stages of hematopoietic differentiation./9

Note that the L in HLDA is for “Leukocyte”. All early CD markers were used to distinguish white cells & their precursors. Later markers expanded to other blood cells (eg CD41/CD61 are on platelets, CD235a on RBCs, CD146 on endothelial cells), and beyond (CD326 is epithelial.)/10

Epithelial cell markers continue to be the Wild West, however. This is why us hematologists get confused when pathologists at patient conferences says things like, “The tumor stained for GATA3, p63, CK5/6, and p40 but dim S100P, variable CK903 - we didn’t do UPII" - um, sure. 🤪

This issue of the Journal of Immunology @J_Immunol from 2015 focused on the CD nomenclature. Nota bene: Although I've been living in the CD world for 20+ years and did lots of "homework" for this tweetorial, I'm not an immunologist, so apologies for any errors.../12

"Human Cell Differentiation Molecules" (#HCDM) is the organization which runs HLDA Workshops. They have a formal process for submitting proposed additions to the CD table (last round closed 12/19). Website: hcdm.org Current chair is @pablo_engel @UniBarcelona./13

I'm just going to go on record right now as saying that if #CD420 is not used for a cannabinoid receptor, it will be a missed opportunity.😄/14

If a molecule has not been well characterized, or has only one mAb at the time of a workshop, it is given a provisional indicator "w" - as in "CDw186" which eventually became CD186, C-X-C Chemokine Receptor 6. I was once told w stood for “weak”, but it just means “workshop”./15

TBH it is not ideal to have CD1a, CD1b, CD1c, CD1d, CD1e - or worse, CD3d, CD3e & CD3g without an a,b,c, or f. (That reflects CD3δ,CD3ε & CD3γ chains, which are part of the T cell receptor.) But biology and terminology is nothing if not messy. We still use the term "malaria"!/16

Mouse CD markers are well-defined. A nice catalogue of mouse CD markers is at bdbiosciences.com/go/mousecdmark…. (I have no @BDBiosciences COI.) Other species do not have their own CD catalogues yet. But CDs can be conserved across species, eg rhesus monkeys overlap a lot with humans./17

I won't dive into CD/antibody therapeutics, but CD20 alone is targeted by 6 FDA-approved antibodies: rituximab, obinutuzumab, ofatumumab, ocrelizumab and radiolabeled ibritumomab tiuxetan & tositumomab – as well as some in development, eg ublituximab (TG-1101). Image: @nejm/18

Of course all of these molecules/antigens have physiologic purposes – they aren’t there only to help us cell sort with #FACS or characterize the origin of malignant cells.🙂/19

Nowadays new molecules/cell-specific antigens are now generally first identified by molecular genetics or proteomics (Image: @broadinstitute), and the antibodies are made *afterwards*. This influences the work of the HLDA Workshops./20

I mentioned in a tweet a while back that to remember which locker I use at my local gym, I restrict locker choice based on CD markers for hematopoietic cells (locker #34, 45, 117 etc). Turns out I'm not alone – many of you share this geekiness. 😜Will we ever use gyms again?!/End