Breeding for Late Blight Resistant Tomatoes

[Joseph Lofthouse]

When I toured Julia's garden, there was a tomato plant that had black decaying leaves. We presumed that late blight killed it. I judged it to be an awesome plant, because it produced dozens of lovely ripe fruits, in a cold, damp growing season near the ocean, where tomatoes presumably don't grow.

I don't have a DNA lab to discern what genes the tomato had, or what killed it. It easily met the adequate for our needs test.

[reed]

I judged it to be an awesome plant, because it produced dozens of lovely ripe fruits

I don't know what else a person could ask for.

I sure wish I had kept up with Captain Crunch. It had awful disease issues, more so than I generally tolerate, but made lots of yummy tomatoes anyway. That was in 2015 and 16 and I didn't really understand and appreciate what I was looking at back then. I may, fingers crossed, still have it archived in the vault.

[Justin Jardinier]

Instead of looking for specific alleles that given a great deal of protection against a very specific disease organism, the idea of horizontal resistance (it seems to me) is to look for genes that make the plant generally stronger, more resistant to any infection, and better able to survive the impact of any disease.

Is there much difference in genes and alleles?

In the case of the self-incompatibility trait in tomatoes, the s-gene has about 35 alleles, meaning that there are about 35 variations of one gene.

Well, in this context of LB, if I am understanding it right, a gene is one position on a chromosome which can be one (or is it the pair?) allele, but we have options for which allele. So the PH genes would be allele options for a specific gene, i.e. that a tomato could be generally very healthy and vigorous, doing well in its environment, like all of my peruvianums for example, or many very healthy domestic tomatoes, or Wildings for example; but, has alleles for that gene which give it no LB resistance.

So, if my understanding of that is correct, then going back to Cathy's point of searching for "genes that make the plant generally stronger", my point is that we do have tomatoes that are very 'strong', whether domestic or wild. But, very weak when it comes to LB. So an example of a 'gene' that could make these tomatoes very 'strong' in the face of LB, is that very gene which the PH alleles are options for. So that was my point when asking "Is there much difference in genes and alleles?"

Wouldn't this be so in general? I mean, unless you are actually adding to the chromosome, adding new genes, like lengthening the chromosome (and from what I understand we can actually rearrange the chromosomes via some interspecial crossing but that may be a different topic), then it would seem to me that looking for strong 'genes' is synonymous with looking for strong 'alleles'. Is it not? And I don't mean to say that this necessarily has to be looking for single alleles. For example, there might be a number of options of alleles for one gene, like with the s-gene, where there might be many options which give no LB resistance, but with several that do. In which case, one could have an (entirely or partially) outcrossing population in which one may have eliminated all non-useful alleles but included a number of useful alleles, for a particular gene, and thus be a well protected population.

I don't know how the resistance works in LA1777. From what I gather it is a number of genes, and I dont really understand what QLT means but is it perhaps that 1 QTL is composed of a number of genes, so it sounds to me like perhaps it involved quite a few different genes in total. Now I don't know if the members of the population have variations of those, all of which are resistant, or whether they are perhaps all homozygous for all of those genes, such that they all have the same type of resistance, unchanging. I could imagine both possibilities.

I would rather gather together 100 alleles that contribute 1% resistance. That explains why I am not working on stacking resistance genes into a single variety.

Isn't that a kind of stacking? Just choosing a larger number of weaker genes to stack, as opposed to a smaller number of stronger genes to stack?

I do also see the potential benefit of having overall resistance caused by the former method. This is also why I intend to do my best at making some populations which do not have the PH genes. Though I will not be able to know what's really going on, my aim would be to allow for the potential coming together or varied forms of resistance in a manner which seems aligned with what you have expressed.

I'm not going to discount his work by approaching it with black/white thinking.

I don't think I would categorise my own position as being founded on black and white thinking. The apparent inconsistencies in his logic do certainly play a role in my not accepting his claims as true, and that might seem black and white. But his total lack of evidence or examples in what I read of him explaining that, were also very significant for me. And I can't call that 'black/white thinking'. I just find it hard to take on faith wonderful sounding things that are not grounded in evidence. (And when put together, numerous logical inconsistencies combined with the absence of evidence, is for me a pretty bad combination). Perhaps he does have evidence, and just did not mention any in that chapter and the rest of what I read. Anyway, I can't accept his logic if it seemed self-contradictory, but I remain open to that which he wrote that seems theoretically plausible at least, but in a skeptical way, waiting for evidence until I would take it on.

It also occurs to me that anyway when it comes to 'horizontal' vs 'vertical', these are after all concepts, attempts at making conceptual classifications and imposing them on the non-conceptual real world. We humans do that, it is a good adaptive function. And in this case the concepts have been constructed for the purpose of prediction and manipulation, such that we can intellectually map breeding phenomena and use them to manipulate the plants to our breeding needs or aims. So I am not too worried about the concepts, so long as my work with the plants I am doing goes ok. And the idea of breeding for LB without the PH genes probably fits with the horizontal conceptual framework, as well as fitting with just a general plan of varied resistance that might be more sustainable in the long run than the specific PH genes. So in the end it might not matter so much which concepts one subscribes to, if in practice the same actions are taken.

And it seems that Ph-2 has been broken. Some strains have broken Ph-3 too.

From the studies I saw, from what I remember it seemed PH-2 is weaker than PH-3 but still useful. I guess it may depend on which strains of LB but I have seen studies where PH-2 did better than no PH; PH-3 did better than PH-2; but PH-2 + PH-3 did better than just PH-2.

I also remember hearing that although PH-1 might be generally considered not useful anymore, that adding that (I guess to other PH combinations) can help too, somewhat. Now I will try to find a paper showing the PH 2 and 3 tests....
Ah ok so I found that PH-1 rerefence while searching. In the PhD dissertation 'Late blight (Phytophthora infestans) on tomato- evaluation of pathogen population structure in Britain and development of resistant tomato cultivars for growing outdoors', Stroud, James, 2015:

Furthermore, the experiments presented in Chapter 4 provided some evidence of a residual effectiveness of the Ph-1 gene, at least in providing a slight reduction in infection efficiency by some isolates. Given the "synergistic" effect of combining the Ph-2 and Ph-3 genes, it seems that an attempt to combine Ph-1 with other resistance genes (perhaps by crossing an OP cultivar carrying Ph-1, such as New Yorker, to lines with Ph-2 and/or Ph-3) would be instructive, especially if they were tested against a broad range of P

Here are a couple of charts from the LB trial in that. I added in red what PH genes they have:




There was another study I saw somewhere, plus Roland's ... was it Roland? Somebody here I think... who did a very similar trial, where there were given images of all the young plants, and it was really easy to see the differences in the resistances given by the different combinations of PH-2 and 3. But here anyway is a chart giving written info, which confirms PH-2 + PH-3 combined giving better resistance than alone, and both homozygous better than heterozygous. And this page was updated on 2022, though not sure if the data might actually be from 2014:
https://eorganic.org/node/10822

And ah I found Roland's report, which was in this very thread! This shows to me that PH-2 is still very valuable:

The status was drawn up 5 days after the 2nd infection:
90% affected on leaves and stem: (no resistance genes)
75% affected on leaves and stem: (ph2 homozygous)
75% infected on leaves, stems are free of infection: (ph3 heterozygous)
50% infected on leaves, stems are free of infection: (ph2 and ph3 heterozygous)
<10% affected on leaf: (ph2 and ph3 homozygous)


For pictures check here:
https://www.facebook.com/permalink.php?story_fbid=pfbid02ARvTSK8GhHgzADrKFXqNZcBgvsvvAkFRj32mgbN7HXeFppFRy8QJrggp4WwmS95Kl&id=100024953531011

From my understanding on US studies, it is not possible to deduce Ph-X genes from resistant plants, it could be Ph-3/Ph-3 or Ph-3/+ Ph-2/+ or stronger combinations. Could be different in EU depending on strains.

To me, the above trials imply that we can potentially breed and identify PH-2+3 homozygous plants either by selecting them at the end of the season via natural exposure to LB, or perhaps preferably, when young as Roland did, visually. I suppose it might help if we deliberately included some know to be homozygous for both, and some lesser ones, for comparison purposes. And perhaps it might depend on what LB strains we are infecting them with. But anyway this method seems like it could work really quite well. And even if it didn't work perfectly for the first round of selection, perhaps letting some homozygous PH-2 but heterozygous for PH-3 through or vice versa (Roland's results would suggest those are the only ones not homozygous for both that might slip through), perhaps that can be sorted out the next year/s through further selection, which might go differently due to differing LB strains.

Or, this method could be used to do the vast majority of selection and then use marker selection for those remaining, which should massively reduce the cost of marker assisted selection.

Yes LA1777 is stronger, it is each individual resistance genes that are weaker that any Ph-X. LA1777 has at least 5 QTLs, but i've seen no data about each QTL power.

Yeah that seems pretty tricky working with that. Plus some may be associated with negative traits also, apparently.

I just came across this summary of the LB resistance genes which others might find pretty useful, from the abstract of the paper 'Fine Mapping of the Ph-2 Gene Conferring Resistance to Late Blight (Phytophthora infestans) in Tomato' - Xiaona Zhi et al, 2021:

Breeding new varieties with genetic resistance to P. infestans is the most effective way to minimize late-blight-caused losses. Thus, the effective use of new resistance (R) genes by breeders is essential. The resistance to P. infestans is mediated by R genes and several quantitative trait loci (QTLs). Over several decades, researchers have identified late blight R genes, including Ph-1, Ph-2, Ph-3, Ph-4, and Ph-5, in wild tomato species. The Ph-1 gene, which confers complete resistance to P. infestans race T0, was mapped to the end of chromosome 7 (Bonde and Murphy 1952; Peirce 1971). The Ph-2 gene, which was identified in S. pimpinellifolium 'West Virginia 700' (Gallegly and Marvel 1955), confers partial late blight resistance, which is useful for slowing the spread of the disease (Black et al. 1996a; Goodwin et al. 1995). The Ph-2 gene was mapped to an 8.4-centimorgan (cM) region between markers CP105 and TG233 on chromosome 10 (Moreau et al. 1998). The Ph-3 gene was identified in S. pimpinellifolium 'LA3708' (AVRDC 1994) and is a partially dominant gene located on the long arm of chromosome 9, where it encodes a coiled-coil nucleotide binding site leucine-rich repeat (CC-NBS-LRR) R protein (Black et al. 1996a,b; Chunwongse et al. 2002; Zhang et al. 2013, 2014). The Ph-4 gene was detected in S. habrochaites 'LA1033' and was reported as a potential QTL (AVRDC 1998; Lough 2003). The Ph-5 gene was identified in S. pimpinellifolium accession PI 270443, which is resistant to at least seven late blight pathogen races (Foolad et al. 2006). However, Ph-5 is controlled by two sites, one on chromosome 1 (named Ph-5-1) and one on chromosome 10 (named Ph-5-2) (Merk and Foolad 2012; Merk et al. 2012; Nowicki et al. 2012), which has increased the difficulty of its application. Additionally, some late blight resistance QTLs have been identified in S. habrochaites (Brouwer et al. 2004; Chen et al. 2008; Li et al. 2011), S. cheesmaniae (Lukyanenko 1991), and S. pennellii (Smart et al. 2007). Unfortunately, introgressing these disease resistance QTLs into cultivated tomato plants requires more time, and many of them may be associated with undesirable traits (Brouwer et al. 2004). To date, the QTLs conferring late blight resistance have not been successfully applied to commercial production, and their practical value remains unknown (Foolad et al. 2014).

I was not aware of the part I highlighted. That's especially interesting about cheesmaniae - I made quite a few cheesmaniae crosses, so I will keep an eye out for resistances in their progeny.

Uniformity in my view is what makes resistances quickly obsolete. Large fields of the same crop variety.

That's what makes me afraid of relying too much on the PH genes - their widespread public use. But on that note, I ponder the significance of small scale diversity vs large scale. By which I mean, might diversity at the individual population level (in one garden) be more significant for a pathogen transmitted from plant to plant over short distances, but relatively less significant for pathogens spread by the wind over long distances (like LB)? By which I mean, perhaps for the latter, the 'population' from the perspective of the pathogen, might be the actual large scale population of tomatoes throughout an entire region.

So if there is not diversity in that overall region, perhaps we are at a lesser advantage than we might be, with our own small scale diverse population?

To give an example - if we have some PH-2 and some PH-3 plants in our own population, but if the pathogen takes advantage of the huge breeding medium of a region's tomatoes, then it may overcome the entire region's PH-2, including our own. In which case our own PH-2 may have been at no advantage to those in homogenous populations. Similarly, if we have all the combinations of PH-2 and PH-3, each individual may fair as well/poorly as all those combinations as they exist in homogenous populations within the region.

This makes me think that having tomatoes, whether diverse or homogenous, that have either 'unpopular' forms of resistance, or that + the PH genes, might do better. Given that unpopular ones, even if grown homogeneously on the small scale, would provide a smaller medium for evolution of the pathogen. And in terms of scale, might be equivalent to individual plants in a varied population for a pathogen with shorter reach. I'm just speculating here but anyway those are my thoughts so far. Either way I prefer (in addition to PH based populations) the idea of a diverse outcrossing or partly outcrossing population with unpopular resistance anyway since I want to breed also for other traits, so I would need to allow that LB resistance to cross enough to get as good as the source, or hopefully better by including various different sources, and maybe be able to continue to adapt as LB itself adapts.

[William Schlegel]

One possible approach to LB breeding would be to breed for speed to production of ripe fruit. If LB hits during a certain portion of the growing season. You might be able to do late blight avoidance just like drought or heat avoidance. Tomatoes that have already produced enough usable fruit before the late blight kills them. Same with other tomato disease. Speed might be an alternative to Resistance.

Julia, for example, might not need resistance at all, if the tomatoes can produce enough tasty fruit in her cool growing conditions, before the late blight hits.

One of my family members described the tomato plants melting with the fall rains in the Seattle area. That shift in weather leads to a huge problem. One approach might be to only grow tomato plants that produce well before that climactic wall comes crashing down.

I have a climactic window for premium fruit in the summer dry season. I can stop watering entirely, and ripen extra tasty crack free tomatoes. Rains and I get bad cracking, spoilage at the crack sites, if the cracks heal nicely they still break open to transport. I also get reduced flavor quality as growing conditions get colder and colder. Problem! So that mid summer fruit is best quality. A lot of the varieties reknowned for best flavor can't produce that flavor under my conditions because they overshoot my best flavor window. It's far more likely for disease to be a problem also after the best flavor window.

Black cherry is a candidate for premium fruit after the flavor window for me. Tastes good cold grown. It might be worth crossing into PHX lines for Julia and my Seattle relatives.

By adding speedy varieties that can out run late blight into the garden mix it would tend to satisfy my concern about uniformity of genetics too.

Weird maybe botanists only thought: why not grow some of the hard to breed with species tomatoes just for their own intrinsic value as cool plants. Then they are present to mess with the pathogens by adding diversity and we just don't have to expect things of them? At least not in the short term- like our personal life times. Also they might feed a more diverse array of solanum pollinating bees. Ooh- I noticed in a South America field study report I shared recently that some of the bee species buzz pollinating wild tomatoes in South America are small species. Our North American bumble bees Bombus are large species.

[Cathy A]

"My point is that we do have tomatoes that are very 'strong', whether domestic or wild. But, very weak when it comes to LB. So an example of a 'gene' that could make these tomatoes very 'strong' in the face of LB, is that very gene which the PH alleles are options for ."

The problem with this approach is that a single mutation of the disease organism can wipe out the resistance provided by that single gene.

The point of 100 genes each of which adds 1% resistance is that no single mutation, or even several mutations, can knock out all of the resistance and lose the crop in a single generation.

[Cathy A]

William, that's pretty much what I have done in past years. I grow early varieties that produce plenty of fruit before being killed by disease.

Some years disease affects only a few plants. This year was my worst disease year ever, but since it was a record flood disaster year with FEMA opening up local offices, it was not a normal growing year.

The dense tomato planting I used for trials may also have made disease spread faster.

[Justin Jardinier]

The problem with this approach is that a single mutation of the disease organism can wipe out the resistance provided by that single gene.

The point of 100 genes each of which adds 1% resistance is that no single mutation, or even several mutations, can knock out all of the resistance and lose the crop in a single generation.

Yes, in theory that sounds great. My issue with it is that it would take a lot more work to stack 100 genes than to stack 2 or 3. And I just don't know of any edible tomatoes that can put up with the conditions in some regions where LB is particularly serious. So, since I do want to eat tomatoes, and do want to make tomatoes that others can eat also, I am trying to go for the best that is available, but trying multiple methods, so that I hope to have multiple resistant populations, some of which should be likely to outlast the others. I.e., not putting all my eggs in one basket. Also I would expect the non-PH populations to do not as well (at first anyway) and take longer to improve or take relative first place as LB adapts, so I intend to at least have some PH-based populations to give good food in the meantime. Also PH-5 not being in commercial tomatoes yet (so far as I understand), might give PH-5 populations a nice advantage for the moment. So having PH-2+3 homozygous populations (would be lovely to add PH-1 to that too though I'm not planning on that) as well as PH-5 and non-PH populations, seems to me the safest bet.

Plus, even if working with resistance from non-PH based varieties/populations, without genetic testing, we might never know if our resistance is coming from 100 genes, or 7 genes, or 2 genes. So perhaps it is also quite hard to know whether one is using that method of a vast number of resistance genes, even if holding that intention.

[ThomatoGarden]

Hello all,

i was able to select 4 Mountain Magic F2 Plants with as good Resistance as Primabella. (2022 was no blight)

One had even bigger fruits (5cm) . Blight was really strong this year even one Primabella Plant had little Problems!

Maybe a selection of Mountain Magic would be a good base for crossing Blight Resistant varieties.

Primabella was created through selecting the best plants from Philovita F1 and trialed at different Places over many years under field conditions.
Philovita F1 is nowadays not very Resistant at all.

So Mountain Magic has a much higher future potential than Primabella cause it's allready very resistant atm.

Next year i will trial the F3 Generation of the big ones.


Greetings Thomas

[reed]

Wow, a lot to go back over in these last several posts. A few things pop out as specifically related to my garden. Concerning the fast maturity versus resistance, I'd like both. The big sweet bicolors that I like are pretty long season, no way to beat any diseases for them, they have to endure the full range.

To the point of diversity within the population, that's where I think mine come up pretty short. I want to start mostly with those I already have and try just crossing them with each other and growing others to evaluate for adding in later. I'm not good at all with hand pollinating and I got distracted by something shiny and fell behind with tomatoes. Here are some old photos of what I have to work with, next year. I have no idea why I have so many with open flowers.

Concerning insect pollination, my bumblebees are not all that fond of tomatoes until very late in the season when a lot of other flowers are done. Several smaller bees and flies like them a lot, all season.   

[Joseph Lofthouse]

My pepo winter squash commonly get killed by powdery mildew. I feel fine about that, because it happens very late into cold, damp fall, after they produced an abundant harvest.

[Joseph Lofthouse]

The reason that i intensely focus only on 100% self-incompatibility in my tomatoes, is that the plants themselves can do the gene-stacking of all those minor QTLs.

[William Schlegel]

That is some decent exsertion of the style and stigma. Might be worth assaying some of your saved potato leaf seed for existing outcrossing rate.

With the crosses I managed to make and send you with your Utah Heart, Hoosier Rose, and Mr. Stripey. Alot of the usable diversity that might segregate out of those is in the earliness or other categories. In 3/4 of the crosses the resistance to your growing conditions should come from your genetics not any I added! Unless the earliness really helps!

Utah Heart x (MMM x Sweet Cherriette)
This cross is 1/4 the shortest season cherry tomato I know of. 1/8th Joseph's Big Hill. 1/8 my Mission Mountain Sunrise which has the blue from Blue Gold and is really short season likely from Joseph's Brad. So small chance of getting blue skin from the Blue Gold ancestor, Small chance of getting a nice open flower from Big Hill, but quite a good chance of picking up some serious earliness.

Utah Heart x (MMR Dwarf)
Good chance of getting yellow fruit, good chance of getting dwarf. Small chance of open flowers with exserted stigma from the Big Hill ancestor. Though Utah Heart might not be too bad for that from the photo above. Small chance of early.

Mr. Stripey x (MMM x Purple Zebra F1)
Small chance of blue, Small chance of open flower, small chance of earliness, small chance of stripes, small chance of resistance, 50% chance of bicolor on the fathers side (you might see some F1 bicolors from that)

Hoosier Rose x (MMM x LA1375)
Small chance of blue, Small chance of bicolor, Small chance of brown rugose fruit virus tolerance. Small chance of earliness.

I would expect that the portion of seed I retained from those crosses, should eventually adapt to my garden and be less and less likely to thrive in yours- especially on past the F2 in the F3 and subsequent unless I saved a lot of seed I wouldn't normally save to retain as much diversity as possible and refrained from finding my favorites and just saving seed from them.

[Cathy A]

My issue...is that it would take a lot more work to stack 100 genes than to stack 2 or 3...[I am] trying multiple methods, so that I hope to have multiple resistant populations, some of which should be likely to outlast the others. I.e., not putting all my eggs in one basket.

I agree, but I do not consider utilizing a handful of PH genes to be "multiple methods."

While it's more work to stack the 100 genes up front, given a growing area with a steady background of disease pressure and minimal selection, I wouldn't expect it to be difficult to maintain once created.

Stacking a handful of carefully-selected genes may be less work up front, but you will never truly be caught up, because a single new disease mutation can nullify one or even all of these genes. And you could lose most or all of the crop the first year the new disease mutation reaches your garden. Then it will take multiple years to find a new single-gene with resistance and cross it into your existing varieties.

Joseph is thinking along the same lines I am.  A landrace, rather than a single variety, with the plants doing most of the crossing on their own. Eventually I hope we'll have one or more solid landraces of tomatoes that are largely or entirely SI and also produce 100% edible, tasty fruit. Once we reach that point, maintaining them should be pretty easy.

Tomatoes are one of the hardest food plants on which to practice this method, because they are highly prone to disease AND they are mostly self-pollinated. That doesn't make the method less desirable, just harder to do.

[William Schlegel]

The reason that i intensely focus only on 100% self-incompatibility in my tomatoes, is that the plants themselves can do the gene-stacking of all those minor QTLs.

I think a 50 percent SI tomato population would also be intensely interesting. I may have a mixed SI and SC system in my garden just by virtue of my interest in the wild SI species, but they and their early generation hybrids are stuck in wild type unpalatable tomato mode in the short term.

Though I think quite a bit of gene stacking could also happen from any segregating population of tomatoes even from hand crosses.

[Roland]

Hello all,

i was able to select 4 Mountain Magic F2 Plants with as good Resistance as Primabella. (2022 was no blight)

One had even bigger fruits (5cm) . Blight was really strong this year even one Primabella Plant had little Problems!

Maybe a selection of Mountain Magic would be a good base for crossing Blight Resistant varieties.

Primabella was created through selecting the best plants from Philovita F1 and trialed at different Places over many years under field conditions.
Philovita F1 is nowadays not very Resistant at all.

So Mountain Magic has a much higher future potential than Primabella cause it's allready very resistant atm.

Next year i will trial the F3 Generation of the big ones.


Greetings Thomas

Primabella is currently the most resistant variety available in Europe. It is an inbred line of Philovita F1 but has been crossed in the field by an insect of an unknown variety.

Philovita F1 is heterozygous for ph-3 and Primabella is homozygous for ph-2 and ph-3.

Mountain Magic F1 comes from a cross between a resistant beef variety (NC 2 CELBR) x a sweet cherry tomato (NC 2 grape). And is heterozygous for ph-2 and ph-3 genes. He is also heterozygous for the rin gene for extra long fruitlive and not ripening fruit. And there is the br gene (brachytic) what means the fruit clusters are closer and need les internodes.

How many F2 plants have you grown?
It is certainly possible that you have found some very resistant selections in F2 that are equivalent to Primabella's resistance.
In many tests, the resistance of Mountain Magic F1 appears to be very high.

Have you also noticed differences in taste in your F2 plants and perhaps you have found the rin gene that do not ripen?
Or did you find the br gene that made the plants more compact and the fruit clusters were every 2 leaves instead of 3 leaves?

ps. you can also find determinate plants in the F2.

I think Mountain Magic have a lot of potential to create nice OP varieties.