So what’s all this business about A2 milk?  How is it different from A1 milk?  Is it hooey?  What do these codes refer to anyway?

Starting with the last question first, A1 and A2 refer to “versions” of the beta-casein gene. In this case, the gene in question encodes the protein beta-casein, one of three casein proteins, which is, of course, a key component of cheese. The A2 version of this protein varies just a little bit in its structure from the A1 version. An individual cow’s genotype could be A1A1, A2A2, or A1A2 for the beta-casein gene; her milk would then contain whichever protein versions her genes dictate.

The a2 Milk Company claims that A2-only milk is easier on digestion. A recent study has suggested that in milk-sensitive individuals, milk containing A1 protein may be associated with symptoms of discomfort after milk consumption, and with strictly A2 milk those symptoms may be lessened. Also, gastrointestinal transit time appears to be slower with A1 milk consumption but, contrarily, yielding softer stools. Additional studies to provide replication of these findings are needed. Early studies on A1 and A2 milk that suggested a link between A1 milk and several diseases have been unsupported by subsequent investigations. That is, there is no evidence that consuming milk containing A1 protein carries any disease risk.

So, should breeding decisions be made on the basis of beta-casein genotype? Or perhaps the question is, is A2 milk a fad or a legitimate, long-term slice of the market? Does A1 vs. A2 matter for yogurt or cheese making? There is a lot we still don’t know about the biology of milk. However, there would seem to be little risk in choosing A2A2 bulls. There are quite of few of them out there, although the number varies by breed. The most comprehensive and recent documentation of beta-casein genotype by breed has been compiled by the Canadian Dairy Network (see table).

Available data from U.S. cattle are more limited in number and over 20 years old. Interestingly, Zebu- or Brahman-type cattle have a very high frequency of A2A2.

Bottom line, don’t compromise your primary genetic objectives for your herd just to chase an A2A2 genotype, but there’s likely no harm in moving that direction if it makes sense for your market of the future.

For a deeper dive into the biology of A1 and A2, read on.

A1 and A2 refer to types of gene variants of the beta-casein gene. A gene variant (an allele, for those who remember their biology) is when we have a difference in the DNA sequence for a particular gene. The A2 gene variant encodes a proline (a particular amino acid; you’ll recall that amino acids are the building blocks of proteins) at position 67 in the 209-amino acid chain that forms beta-casein (see figure below). The A1 gene variant encodes a histidine at position 67.

The A1 beta-casein protein is thought likely to be cleaved (cut) during gastrointestinal digestion at the position 67 histidine, while A2 beta-caseins are less likely to be cleaved there. Cleavage at amino acid 67 generates a short protein (a peptide) called beta-casomorphin-7 (abbreviated BCM7). BCM7 has opioid properties. Now, no need for concern. We all know that while milk is tasty, it is not a very effective painkiller nor brain manipulator. It is possible though, that BCM7 may have some effect on processes in the gut, such as slowing the rate of passage. Also, all milk contains additional types of opioid peptides. Other foods (from animals and plants) do as well.


Diagram shows cartoon of regions of beta-casein protein with variable amino acid indicated.
The region of the beta-casein protein where A1 and A2 vary.

When heifers calve very young, there is a greater risk of stillbirth and lower first-lactation milk production. When heifers are old at calving, their fertility may be negatively affected and it raises their culling risk. Plus, there is the cost of feeding them to that age before you get any return. So what is the sweet age for first calving to maximize average lifetime production? To answer that question, researchers at USDA analyzed production, reproduction, and lifetime data along with genetic (relationship) data from 13.9 million Holstein, 1.2 million Jersey, and 90,400 Brown Swiss cows. (Isn’t the national dairy database great? That’s just cows who first calved from 1997 through 2015!) Genomic data from aba Jersey cow andnewborn calfout 205,000 of those animals were also used.

One of the first interesting results of this study was documentation of the significant trend toward younger ages at first calving (see Table 1). It’s been most pronounced for Jerseys.

Table 1. Percentages falling into each age-at-first-calving (AFC) category in 1997 and 2012. (Data condensed from Hutchison et al. 2017.)

AFC (months) Holstein

1997           2012


1997        2012

Brown Swiss

1997        2012

18–22     7.9   33.5   18.5   65.2     3.0   12.8
23–27   66.8   58.3   64.2   31.1   53.5   59.2
28–35   25.3     8.2   17.3     3.7   43.5   28.0

Age at first calving may serve as an indirect indicator of general productivity and survivability, as lower ages at first calving correlate with higher lifetime production and fertility. That is, heifers capable of getting pregnant at younger ages may just be more robust animals in general. In order to capitalize on those individuals, one shouldn’t start breeding too late. The data support a target age of 21-22 months for Holsteins and Brown Swiss to deliver their first calves and 20-21 months for Jerseys. However, breeding at ages younger than 11-13 months is not recommended because younger heifers are more likely to have stillborn calves. The authors of the study suggest that AFC be incorporated in bull selection indexes, which would enable population-level selection for an AFC that increases profitability.

the article: Hutchison et al. Genomic evaluation of age at first calving. Journal of Dairy Science. August 2017. 100:6853–6861.

two hornless calves on mowed grass
Hornless calves (named Buri and Spotigy) produced by gene editing. Photo from Carlson et al. 2016 Nature Biotechnology 34:479-481

We know that hornless cattle are safer for people, their herdmates, and themselves. Unfortunately, the combination of polledness and elite genes for other, more critical traits (like milk yield and productive life) don’t often appear in the same animals. We could spend several decades using polled sires to introgress the POLLED allele (allele = version of a gene) into the broader dairy population, but we would sacrifice gains in other traits, because along with the POLLED allele, the calves would get other stretches of less desirable DNA. (However, the nice thing about the POLLED allele is that it’s dominant, meaning that only one POLLED allele is required. At that same location on the other paired chromosome, there can be the horned allele, but we would still have a polled cow.)

You may have heard of gene editing, particularly with a system called CRISPRs. These CRISPR molecules can be introduced into target cells and are capable of recognizing a particular stretch of DNA and cutting at that location. If pieces of DNA containing the desired sequence for that location (e.g., the POLLED sequence) are made available to the cell at the same time, the cell’s DNA repair machinery will use that “new” DNA to repair the break in the chromosome. Voila! That repaired chromosome now contains the DNA sequence we want at that location, instead of the sequence that was originally there.

This type of gene editing has been successfully done in cattle embryos. In this case, the researchers used a more primitive version of CRISPRs called TALENs, but they do the same thing. In embryonic cells from horned cattle, the targeted section of DNA on chromosome 1 was replaced with the POLLED DNA sequence. In this proof-of-concept experiment, clones were created from these cells. And they grew no horns! The rest of the DNA in these animals remained the same as it was in the original genetic source, only the horned/polled location was altered. The two bulls produced (pictured above) will be used in breeding experiments to confirm that their offspring will also be polled.

This precise gene editing technique could be used to introduce polledness into elite dairy sires. In one generation, we could nearly eliminate the need to dehorn/disbud calves. That’s assuming there are no regulatory setbacks regarding the gene editing technology. (Ah, the potential sticking point.)

If you’d like some additional explanation accompanied by video of the polled bulls—currently residing at UC Davis—Science Friday has that here. The paper can be found here on page 479 (Carlson et al. 2016 Nature Biotechnology 34:479-481). I’ve glossed over some of the details, so if you’d like any additional explanation, please post a question via the “Leave a reply” link or email me.