An Overview of the Backcross Project

By James E. Seltzer, Ph.D. in June of 2006

The uric-acid stone problem is a principal genetic defect in the Dalmatian breed. Since at least 1938 we have known the inheritance pattern of this defect. It behaves like a simple autosomal recessive. This is the same sort of genetic trait as the one that determines whether a Dal will have black spots or liver spots.

It is essential that we keep in mind that the defective trait we are talking about is the very high urinary uric acid (UUA) concentration in Dals. The relationship between UUA and the actual formation of stones is not linear. It is important to observe that: The mode of inheritance of the uric acid defect in Dalmatians is not in dispute.

First, let's make sure we understand the fundamentals. A good place to start is at the web page of the American College of Veterinary Surgeons, ACVS 


Urolithiasis (urinary stones) is a common condition responsible for lower urinary tract disease in dogs and cats. The formation of bladder stones is associated with precipitation and crystal formation of a variety of minerals (magnesium ammonium phosphate hexahydrate, calcium oxalate, urates, and others).

Causes and Risk Factors

What causes urinary stones? Several factors are responsible for the formation of urinary stones. The understanding of these processes is important for the treatment and prevention of urinary stones. In general, conditions that contribute to stone formation include:

   a high concentration of salts in urine

   retention of these salts and crystals for periods of time in the urinary tract

   an optimal pH that favors salt crystallization

   a scaffold for crystal formation

   a decrease in the body's natural inhibitors of crystal formation. 

The Backcross project is primarily concerned with the first of these since this is directly related to the uniquely Dalmatian genetic defect mentioned above (Trimble, HC, and CE Keeler. 1938. The inheritance of "high uric acid excretion" in dogs. J. Heredity, 29, 280-289.)

The other contributing factors to the urinary stone problem are interesting and worthy of discussion, but we should not confuse these with what the Backcross project is about.

In this overview I will try to explain:

•   How one introduces a normal version of the gene into the Dalmatian breed. 

•   How one identifies and isolates that gene in the progeny. 

•   How one insures that the normal gene is passed on to succeeding generations. 

•   How one validates that the Dalmatian uric acid defect has been corrected. 

•   Current status of the Backcross project and location of the defective gene in the Dalmatian genome. 

•   Alternative approaches for dealing with the uric acid defect (such as pedigree analysis and selective breeding).

How one introduces a normal version of the gene into the Dalmatian breed.

To anticipate and avoid arguments about inviolability of pure breeds and racial purity, I need to say up front that pure canine breeds exist primarily in the minds of the dog fancy and are simply paperwork exercises codified in the registries of the various national kennel clubs. They do not exist in the flesh and blood reality of dogs living in the real world. Dog registries and closed stud books are a recent invention of today's dog fancy - originating only a little more than a century ago. The partnership between man and dog reaches back much further. Robert K. Wayne of the University of California, Los Angeles and his colleagues now have evidence that dogs could have been domesticated 100,000 years ago -- if not earlier.

What comprises a breed is not a unique set of genes, neatly packaged with clear boundaries that identify what is and what is not a member of the breed. AKC registration is not especially meaningful for defining the attributes of the Dalmatian. Even a cursory visit to Sue MacMillan's coat color web pages will quickly shatter such an illusion.

Purebred Dalmatians, presumably AKC registered, can be found in brindle, lemon, orange, blue, tri-color, and sable. Dals share these genes with other pure breeds. In Dals, these alleles are fairly uncommon; in other breeds they are both common and in many cases desired. There is no doubt that genes that control other conformational attributes (e.g., ears, height, tail set, etc.) are also shared with the other so-called pure breeds.

What distinguishes one breed from another is the relative allele frequencies of the aggregate set of genes that serve as blueprints for the breeds of dogs. Dalmatians, for example, have a higher frequency for the extreme white piebald allele (sw) and the ticking allele (T) than the cocker spaniel -- but Dals do not have exclusive ownership of either of these alleles. Dals just have these alleles in greater abundance.

Most breeds of dogs have a normal gene for uric acid excretion, and, compared with Dals, rarely have problems with urate stones. The ancestor to the Dalmatian also had such a normal gene, but that gene got lost in the shuffle as the breed was propagated and artificial selection was taking place. The normal gene may have been closely linked (on the same chromosome) with another gene that was considered a desirable characteristic by the early breeders. On the other hand, the normal gene may simply have become victim to random genetic drift and got lost along the way, which is not unusual when the number of dogs being bred is small. However it occurred, to the best of our knowledge, there were no Dalmatians anywhere that still carried the normal uric acid excretion gene prior to the Backcross project.

Since that normal gene did not exist within the Dal breed, it was not possible to use breeder selection methods to increase the normal allele frequency and thereby diminish the incidence of urate stone disease in Dals. (Which answers the question of why we can't reestablish the normal gene in the same way that we can establish a true-breeding, liver-spotted line of Dals.)

To find the normal gene it was necessary to turn the clock back to the point in time before the Dalmatian breed branched off from its kin at their common origins and followed its own path. We don't know exactly what the common ancestor(s) was at that early branch point, but we can surmise what its progeny probably look like today even though they followed different selection paths during the intervening generations. Considering a broad array of phenotypic attributes, and the likelihood of a not-too-distant common ancestor, Dr. Bob Schaible selected the Pointer as a probable descendent of that closest common ancestor.

When a Dalmatian was mated to a Pointer, all the cross-bred pups carried one copy of the normal uric acid excretion gene that it got from its Pointer sire. Since, according to the early work by Trimble and Keeler, we already know that the uric acid defect is a simple autosomal (not sex-linked) recessive gene, all the first-generation pups excreted normal levels of urinary uric acid (UUA) as was predicted by the autosomal recessive model. The first-generation pups, of course, did not much look like Dalmatians.

In order to refine the line it was necessary to cross-breed back to a purebred Dalmatian, hence the name Backcross project. The second generation pups, although they began to look more like purebred Dalmatians, did not all carry a gene for normal UUA. Only about 1/2 of these pups got the normal gene. The best of those carrying a copy of the normal UUA gene, i.e., those that most closely resembled Dalmatians, were selected for further breeding in the Backcross project.

The process continued to select pups 1. for normal UUA, and 2. for proper Dalmatian conformational attributes. The Backcross project has continued to the point that the latest generation pups are tenth generation descendents of the one original Pointer. The lucky one's still carry that Pointer's genetic bequest: a gene for normal UUA. Most of their other genes are derived from their Dalmatian dam, their Dalmatian grandam, their Dalmatian great-grandam, etc.

These pups are still heterozygous for the normal UUA gene. The decision not to breed a homozygous-normal UUA line (yet) has been intentional and relates to the necessity to avoid a genetic bottleneck and all the concomitant headaches that ensue when a line is closely line-bred.

How one identifies and isolates that [normal UUA] gene in the progeny.

The Backcross project started with a Pointer that had normal uric acid excretion (10-60 mg of uric acid in his urine per day) that was mated to a Dalmatian dam with high uric acid excretion (400-600 mg of uric acid per day in her urine). There is no overlap in these numbers; there is no mistaking one for the other. A veterinary lab technician provided an unlabeled urine sample from the sire and a urine sample from the dam could easily tell you which sample came from the Pointer and which sample came from the Dalmatian. (Canine and Feline Nephrology and Urology, Osborne & Finco, 1995, p824)

As noted above the hereditary pattern for the Dalmatian defect is transmitted as an autosomal recessive. Trimble and Keeler (1938) crossed Dalmatians to Collies, and through subsequent crosses determined that the genetic defect in Dalmatians was an autosomal recessive trait.

When a carrier for the defect (one normal gene and one defective gene) from the Backcross line is mated to a purebred Dalmatian (two defective genes), the expected ratio of carriers to defectives in the resulting litter is 1:1, i.e., we expect approximately 1/2 of the pups to be UUA normal and 1/2 to be UUA defective. This is the distribution of the defect that could be expected by the second generation and for all subsequent generations of puppies.

As early as 1968 a method for screening for abnormal levels of uric acid in humans had been published: J Pediatr. 1968 Oct;73(4):583-92., "Urine uric acid to creatinine ratio--a screening test for inherited disorders of purine metabolism. Phosphoribosyltransferase (PRT) deficiency in X-linked cerebral palsy and in a variant of gout."

Another paper that was published many years after the Backcross project had been initiated questioned the use of the UUA:CR ratio test to estimate the 24-hour total uric acid excretion in healthy Beagles. Am J Vet Res. 1994, 55:472-476, Bartges, JW; CA Osborne; LJ Felice; LK Unger; KA Bird; LA Koebler; M Chen, "Reliability of single urine and serum samples for estimation of 24-hour urinary uric acid excretion in six healthy Beagles."

The authors of the 1994 paper found that some spot samples of urine and creatine taken during the day did not correlate well with the 24-hour UUA excretions, and they attributed that "to differences in urinary uric acid and creatinine excretions after digestion, absorption, and metabolism of the diets."

Yet another paper published in 2004, questioned the use of single 24-hour urinary uric acid excretion measurements in healthy humans since uric acid excretion levels fluctuate widely over even longer periods. Rheumatology 2004 3(12):1541-1545; doi:10.1093/rheumatology/keh379, K.-H. Yu, S.-F. Luo, W.-P. Tsai and Y.-Y. Huang "Intermittent elevation of serum urate and 24-hour urinary uric acid excretion."

The authors of the 2004 paper conclude: "The data presented here demonstrate individual variations in serum urate levels and 24-h urinary uric acid excretions in healthy men with serial measurement. Transient hyperuricaemia and hyperuricosuria are more common than expected, and both transitory and monthly variations are important factors to consider when evaluating the influence of other factors upon serum urate levels and urinary uric acid excretion."

Needless to say, this puts the veterinary clinician who is trying to manage urinary uric acid problems in his patients on the horns of a dilemma. The UUA:CR test, it is claimed, is invalid, because of diurnal fluctuations. The 24-hour urine collections are no good because urinary uric acid excretions are found to vary widely when monthly measurements are compared. Further, these monthly variations are not insignificant: median 623, range 389–1565.

Fortunately, the fluctuations in UUA excretions are of far less concern to the geneticist who is armed with foreknowledge that the pups produced in the Backcross line will segregate into two distinctly different classes according to their levels of UUA excretion. If he can demonstrate that whichever test he uses differentiates between a normal UUA level and a high UUA level, and that the two classes do not overlap, then his objective of matching the pups to the class carrying the normal gene and the class of those that are homozygous for the defective gene is solved.

The Dalmatian Backcross project has used and continues to use the UUA:CR ratio test for puppy classification purposes only. Typical results are given below.

More recent studies have also used UUA:CR ratio tests.[Urology. 2003 Sep;62(3):566-70. Carvalho M, Lulich JP, Osborne CA, Nakagawa Y. "Role of urinary inhibitors of crystallization in uric acid nephrolithiasis: Dalmatian dog model."]

The role of urinary crystallization inhibitors is also discussed below. This is relevant since such inhibitors have been postulated as a reason why, though all Dalmatians excrete high levels of uric acid, not all Dalmatians form urate stones.

How one insures that the normal gene is passed on to succeeding generations.

I have already mentioned the use of various urinary uric acid testing procedures and briefly discussed their weaknesses. I observed that the job of the geneticist working on the Backcross project is considerably easier than that of the veterinary clinician treating stone forming Dalmatians. Nonetheless, the Backcross geneticist must select with a high degree of confidence only those Dals that carry the normal UUA gene for further breeding.

Let us assume that there are 8 puppies produced in a litter where the sire carries one copy the normal UUA gene and the dam is homozygous for the defective UUA gene. The pups should segregate into two classes: a low-UUA class and a high-UUA class, and the most probable split is 4 of each. Of course, getting that exact ratio is not guaranteed. In fact, all 8 pups might fall into the one class or the other -- though that outcome is unlikely (about 4 chances in a thousand for either extreme).

The Backcross breeder will use the computed UUA:CR ratios for each pup in the litter and can plot these values as points along the x-axis on a graph. Examining the plot generally identifies the puppies that belong to each of the two classes, since the human eye has the ability to recognize patterns in data. Further, a statistician can analyze the data using a simple algorithm that defines each class on the basis of a minimum variance. Listed below are the UUA:CR ratios as they were provided to me for one set of Backcross pups, Topper X Twyla litter, Aug, 2005, 8 pups:

UUA:CR ratios (mg/dl uric acid per mg/dl creatinine)









It is not difficult to identify the high and low UUA classes. The class boundaries are readily apparent. 

How one validates that the Dalmatian uric acid defect has been corrected.

To reiterate: the genetic defect being addressed in the Backcross project is the high urinary uric acid excretion which predisposes Dalmatians to urate and uric acid stones.

   - Non-Dalmatian normal uric acid excretion (10-60 mg of uric acid per day)

   - Dalmatian range for uric acid excretion (400-600 mg of uric acid per day)

Data already exist that place urinary uric acid concentrations for the low-UUA Backcross Dalmatian class in the same range as urinary uric acid samples taken from healthy Beagles. Without correcting for diurnal variations (which are influenced by digestion, absorption, and metabolism of their diets), direct comparisons of the uric acid concentration samples for Beagles and the Backcross Dals are impossible using the existing data sets.

In this regard I am referring to quantitative assessments of urinary uric acid excretions. I am not questioning the legitimacy of the UUA:CR ratios as a discriminant used to classify the puppies from the Backcross litters according to whether or not they carry the normal UUA Pointer gene.

Although the 24-hour uric acid excretion values were used by Trimble and Keeler (1938) and have been discussed in Osborne and Finco (1995), they suffer to some degree from the same problem as the UUA:CR sample data. That is they do not reveal peak daily uric acid concentrations, nor do they account for the long-term, monthly fluctuations in uric acid excretion levels. However, from a purely practical standpoint, the 24-hour uric acid excretion values for a set of Backcross Dalmatians can be informative.

Without outlining at this time a detailed protocol, I believe that a limited set of 24-hour tests is appropriate and desirable at some time for validating the Backcross Project.

Current status of the Backcross project and location of the defective gene in the Dalmatian genome.

This topic is the probably the most fascinating of any that I have covered in this paper. First, I want to draw your attention to a very recent publication:

Mammalian Genome. 2006 Apr;17(4):340-5. Epub 2006. Linkage analysis with an interbreed backcross maps Dalmatian hyperuricosuria to CFA03., Safra N, Schaible RH, Bannasch DL. _uids=3059479&dopt=Abstract 

"Dalmatians, like humans, excrete uric acid in their urine. All other dogs and most mammals excrete allantoin, a water-soluble compound that is further along the purine degradation pathway. Excretion of uric acid at high concentrations (hyperuricosuria) predisposes Dalmatians to the formation of urinary urate calculi. Hyperuricosuria (huu) is found in all Dalmatians tested and is inherited as an autosomal recessive trait. A genome scan and linkage analysis performed on a Dalmatian x Pointer interbreed backcross detected a single linked marker, REN153P03, located on CFA03. Haplotype analysis of the region around this marker defined a 3.3-Mb interval flanked by single recombination events. This interval, which contains the huu mutation, is estimated to include 24 genes."

A team of geneticists at UC Davis and Dr. Bob Schaible have successfully narrowed the search for the defective UUA Dalmatian gene to a fairly small region containing only about 24 genes on canine chromosome 3. (Dogs have 39 chromosomes with a total of about 30,000 genes.) So the search for the defective gene is rapidly closing in on its quarry.

A marker on that chromosome, REN153P03, is close enough to the actual gene so that the marker can be used to flag the presence or absence of the normal UUA gene. If you are curious you can look at the marker map for canine chromosome 3 (CFA03) at: guyonmaps_data/cfa03.pdf  You will find the marker, REN153P03, in the lower (magenta) depiction on that map.

A lot of information can be found both on the web and in various publications about the use of markers to aid the breeder in the selection of dogs for breeding and the elimination of hereditary diseases. A good foundational book on the subject is the AKCCHF publication, Future Dog, Breeding for Genetic Soundness, by Patricia J. Wilkie.

The significance of the REN153P03 marker to the Backcross project is that its use allows a breeder to identify Dals that carry the normal UUA gene by DNA analysis. Urinary uric acid excretion tests still remain as a valuable alternative. However, the DNA test has the advantage of distinguishing between a carrier and a homozygous normal condition.

How well do the DNA tests correlate with the UUA:CR tests? One of the investigators at UC Davis states: "We have never encountered a discrepancy between our [DNA] molecular testing and these phenotypes (low or high uric acid/creatinine)." So the correlation to date is perfect. Markers do not always give perfect correlations unless they lie very close to the defective gene, so this is very encouraging.

Summing up the progress on the project:

•   The normal UUA gene from the Pointer has been successfully integrated into the genome of the Dalmatian Backcross line. 

•   The Dalmatian defective UUA gene, as expected, behaves like an autosomal recessive gene as reported by Trimble and Keeler in 1938. 

•   The UUA:CR ratio test unambiguously discriminates between puppies that are carriers for the Pointer gene and those that do not carry the gene. 

•   A gene marker has been located on canine chromosome 3 that can be used to identify Dals that carry the normal UUA gene and will also be able to identify Dals that are homozygous for that gene when they are produced. 

•   The gene marker, REN153P03, correlates perfectly with the classifications of Backcross puppies based on the UUA:CR ratio test.

Alternative approaches for dealing with the uric acid defect

In this section. I will attempt here to look at other approaches to attacking the Dalmatian urate stone problem.

1.   Care and management

This approach essentially maintains the status quo. Recommended methods for minimizing the risk for urinary stones include (1) adequate hydration (there is some evidence that bottled, especially distilled, water can be beneficial), (2) provide ample opportunities for urination, (3) limit purine intake, (3) use pH test strips to monitor urine acidity, (4) use allopurinol under veterinary supervision to prevent recurrence.

These basic Dalmatian husbandry procedures can help, but the underlying hereditary defect of high UUA excretion remains. Periods of stress, especially when accompanied by chronic diarrhea, can lead to acidification of urine and the formation of uric acid stones.

Further, unless uric acid levels are carefully monitored, treatment of stone-forming Dals with allopurinol can result in the formation of xanthine stones.

2.   Gene implantation

Now that the gene that causes the high UUA excretion in Dals is in the boresight of the researchers, some have suggested that high-tech gene splicing should be able to solve the Dalmatian defect without resorting to a crossbreeding to install the normal gene. Perhaps some day, but not in the foreseeable future.

"Gene transfer can be targeted to somatic (body) or germ (egg and sperm) cells. In somatic gene transfer the recipient's genome is changed, but the change is not passed on to the next generation. In germline gene transfer, the parents' egg and sperm cells are changed with the goal of passing on the changes to their offspring."

Obviously, Dal breeders want to use germline gene transfer since they want the normal gene that is implanted to be passed on to the puppies.

"Germline gene transfer is not being actively investigated, at least in larger animals and humans, although a great deal of discussion is being conducted about its value and desirability..."

There are so many intrinsic technical difficulties and risks associated with germline gene transfer, that the process is unlikely to offer a practical alternative at this time.

3.   Selective breeding

It has been suggested that even though all Dals excrete high levels of uric acid, not all form stones. Therefore factors other than the uric acid defect must be involved. The reasoning continues: If breed lines that consistently produce Dals that do not form stones can be identified and the environmental and subsidiary genetic factors that mitigate the stone problem understood, then selective breeding should produce stone-free Dals. Although appealing at first, there are serious problems with this line of reasoning.

First, we should note that environmental factors, which were briefly mentioned under 1., above, are not refined by selective breeding. If selective breeding is to prove useful, it must deal with identifiable hereditary traits. Such traits should have a reasonably high heritability to be amenable to artificial selection methods. Further, it must be possible to identify dogs that carry the desirable traits during the period of their lives that they are used for breeding. Unfortunately, the trait does not form stones can only be assigned with certainty after the dog's death.

Second, we should recognize, particularly in light of Dr. Susanne Hughes ultrasound study, that the category does not form stones might be better classified as has not yet formed stones.

Nonetheless, recent studies with both Dalmatians and humans have identified a number of substances commonly found in urine that are known to inhibit crystal formation and the growth of urinary stones. One of the most studied is the naturally occurring Tamm-Horsfall protein which is a thick mucous material produced in the kidneys. This protein is also known to provide some protection against bacterial infection in the urinary tract.

Carvalho M., and others at the University of Chicago found that the amounts of Tamm-Horsfall protein (THP), and glycosaminoglycans (GAGs) were lower in Dalmatians that formed stones when compared with Dalmatians that did not form stones. query_hl=7&itool =pubmed_docsum

If it can be demonstrated that the Tam-Horsfall protein gene in Dalmatians has several variants, and if the correlation between these genetic variants and the stone-forming status of Dalmatians is shown to be high, then selective breeding might help to minimize the stone-forming proclivity of Dalmatians in a carefully selected breeding line. 

A cautionary note must be inserted here. Inhibitors of urinary stone formation do not prevent crystals from being formed and growing. These inhibitors only increase the uric acid concentrations that are maintained in solution before saturation occurs and crystals are produced -- that is the reason they are called inhibitors rather than preventatives.

This concludes an abbreviated foray into the rationale for and science of the Backcross project. I confess to being attracted to the idea of finally doing something positive and of lasting value for the genetic health of our popular canine breed. I recall reading articles by  Stephen Budiansky's "The Truth about Dogs" whick was published in The Atlantic Monthly. The authro of that article chided dog breeders for their emphasis on breed purity rather than genetic health. Budiansky recognized that not every breed carries the same defects, and damage that has been done can be undone. (The Atlantic Monthly; July 1999:The Truthe About Dogs - 99.07; Volume 284, No. 1;page 39-53.)

The sheer diversity of dog breeds, and the fact that up until a hundred years ago -- a blink of an eye in terms of evolutionary time scales -- genes flowed freely throughout the global dog population, together imply that we still have ample genetic reserves that can be drawn on to undo any damage recently done.

If we make the effort, I believe we can make a difference.

Dr. Seltzer first shared this information with members of ShowDals, an online list for Dalmatian breeders and owners, in June 2006. It was then turned into a two part article that appeared in the Fall and Winter 2006 issues of the Dalmatian Club of America magazine The Spotter. Dr. Seltzer has been a member of the Dalmatian Club of America since 1976. He bred Dalmatians under the Willowind kennel name.

LUA Dalmatian Selia