Molecular Epidemiology:

1
Molecular Epidemiology its concept and
applications
Yong Luo, Zhonghua Ma and Themis J
Michailides University of California-Davis,
Kearney Agricultural Center
2
Definition (version 1)
Molecular epidemiology provides the tools (both
laboratory and analytical) that have predictive
significance and that epidemiologists can use to
better define the etiology of specific diseases,
and work towards their control. --- R.C.
Andrew Thompson Molecular epidemiology of
infectious diseases 2000. 326p.
3
Definition (version 2)
A science that focuses on the contribution of
potential genetic and environmental risk factors,
identified at the molecular level, to the
etiology, distribution and prevention of disease.

4
Why molecular epidemiology?
In molecular epidemiology, the molecular
techniques are used to study and solve the
epidemiological problems that the traditional
epidemiological methods can not.
5
Main areas of molecular epidemiology

• Pathogen Identification and diagnosis ????
• Quantification of Initial inoculum and infection
  ???
• C. Pathogen temporal development ????????
• D. Pathogen spatial distribution ????????
• E. Pathogen long-distance dispersal and migration
  ?????????
• F. Dynamics of pathogen population structures
  ???????????
• G. Interactions between host resistance and
  pathogen virulence and pathogenicity ????????
• H. Others ??.

6
Basic molecular markers used in molecular
epidemiology

• Individual (Molecular diagnosis)
• 1. Hybridization,
• 2. PCR-based
• a. classical PCR,
• b. nested PCR,
• c. real-time PCR.
• Population
• PCR-based assay RAPD, ISSR (internal simple
  sequence repeat, MP-PCR), AFLP.
• Hybridization RFLP.
• DNA sequence ITS, IGS, Protein genes- ß
  tubulia, EF1a , Elongation factor.

7
(No Transcript)
8
Animated principle of PCR
9

• RAPD (Random amplified polymorphic DNA) Bands
  are generated by a 10-bp Operon primer.
• SSR-PCR (Simple sequence repeats) Bands are
  generated by a primer of simple sequence repeats

Three of the bands in the diagram are RAPD bands,
and there are 5 polymorphic phenotypes.
10
AFLP (amplified fragment length polymorphisms)

1. Genomic DNA is digested with both a restriction
   enzyme that cuts frequently (MseI, 4 bp
   recognition sequence) and one that cuts less
   frequently (EcoRI, 6 bp recognition sequence).
2. The resulting fragments are ligated to
   end-specific adaptor molecules.
3. A preselective PCR amplification is done using
   primers complementary to each of the two adaptor
   sequences, except for the presence of one
   additional base at the 3' end. Which base is
   chosen by the user. Amplification of only 1/16th
   of EcoRI-MseI fragments occurs.

11
RFLP (restriction fragment length polymorphisms)

• Use the restriction endonucleases to recognize
  the specific DNA sequences.
• Hybridize to probe DNA or amplify by PCR.
• Analyze the variation of amplified bands.

12
Real Time PCR
(1) Forward and reverse primers are extended with
Taq polymerase as in a traditional PCR reaction.
A probe with two fluorescent dyes attached
anneals to the gene sequence between the two
primers. (2) As the polymerase extends the
primer, the probe is displaced. (3) An inherent
nuclease activity in the polymerase cleaves the
reporter dye from the probe. (4) After release
of the reporter dye from the quencher, a
fluorescent signal is generated.
13

1. Pathogen Identification and diagnosis ????

• For the pathogens that it is difficult to
  morphologically distinguish from other species,
• For the pathogens to be Identified from soil,
  seed, plant tissue, water, and other bio or abio
  environments.

• Basic steps
• Design and create the species-specific primer(s),
• Test the primers specificity using a broad range
  of species,
• Test the sensitivity of the primer(s),

14
Design species-specific primers based
on Species-specific sequences of internal
transcribed spacer (ITS) Species-specific
sequences of randomly amplified polymorphic DNA
region (RAPD) Species-specific repetitive
sequences
15
An example of Monilinia fructicola (Ma, et al.)
A common band amplified with a primer M13
The primer shows Monilinia fructicola
specification
Sequence and design primers
.
16
A list of selected pathogens, markers, and
authors for pathogen diagnosis
17
B. Quantification of Initial inoculum and
infection ????????
X2 X1 r Ln --------------
t2 t1
XtX0 ert

• Initial Inoculum
• Propagule density in soil, water or plant debris,
• Spore density in the air,
• Latent infections in plant tissues,
• Infected seeds of seed-borne diseases,
• Disease intensity in the original source of the
  long-distance dispersal pathogens.

Molecular techniques can be used to quantify the
initial inoculum of some diseases that the
traditional epidemiological methods can not.
18
Example 2 Determination of spore density of
Monilinia fructicola in the air (Z. Ma, Y. Luo
and T. Michailides)
Spore density in a stone fruit orchard in early
spring can be determined with a spore trap.
19
Principle of the nested-PCR method
1. From the sequence of species-specific
segment of DNA, determine a specific segment.
2. Design the sequences of external and
internal primer pairs. 3. In PCR, amplify
the specific segment of DNA by using the external
primer pair. 4. Use the internal primer
pair to amplify the segment of the PCR product
which was amplified with the external primer pair
previously.
20
Test of sensitivity of nested PCR
EMfFEmfR
IMfFImfR
Internal and external primer pairs
21
Test of real spore trap samples
This method can be used to determine the spore
density of gt300 spores/sample/day. This spore
density may induce 5 blossom infection under a
favorable condition. This 5 blossom infection
can be used as a threshold for fungicide
application.
22
C. Pathogen temporal development ????????

• Studies focus on pathogen population dynamics
  over time.
• Isolates should be collected from different time
  in same or different fields.
• Proper molecular makers should be used to produce
  polymorphic DNA haplotypes or genotypes.
• Analyses will include
• Genetic distances among sampling populations at
  different times.
• Determination of genetic identity and distance
  among sampling populations.
• Determination of changes in genetic structure
  among populations.

23
Example 1
Analysis of Population Structures and
Dynamics of Monilinia fructicola in California by
Using Microsatellite Primed-PCR  Yong Luo,
Zhonghua Ma and Themis J. Michailides  
Department of Plant Pathology, University of
California, Davis, Kearney Agricultural Center

• Isolates were collected from plum and prune
  orchards in 2001.
• Time of collection in the plum orchard
  apothecia, fruit at mid-season and fruit at
  harvest.
• Time of collection in the prune orchard mummies
  in early spring, blossoms, fruit at mid-season,
  and fruit before harvest.

24

• Five microsatellite primers, M13, (AAG)8, (AG)8C,
  (GACA)4, and (AG)8C, were used in the PCR
  amplification.

• Data analyses included
• UPGMA tree development.
• Calculation of genetic diversity for each
  sampling group.
• Calculation of genetic identity and distance
  among sampling groups.

25
Percentage of polymorphonic loci and genetic
diversity for each sampling group
26
Genetic identity (above diagno) and distance
(below diagno) among sampling groups
27
Conclusions

• In the prune orchard, the pathogen populations
  remained a general consistency in genetic
  identity over the growing season with a slight
  change from the beginning to the end of the
  season,.
• This change was related to the changes in the
  genetic diversity within the individual sampling
  groups.
• In the plum orchard, the genetic structures of
  the pathogen populations did not change much over
  the season in the orchard.
• However, the genetic structure of populations
  directly isolated from apothecia was
  comparatively less consistent than those of
  sampling groups in the rest season.

28
Example 2
Spatiotemporal Analysis of Population Structure
of Botryosphaeria dothidea from California
Pistachios Zhonghua Ma, Yong Luo, and Themis J.
Michailides Dept. of Plant Pathology, University
of California, Davis, Kearney Agricultural
Center
29

• 378 isolates were recovered from pistachio in
  seven counties of California from 1997 to 2001.
• Six microsatellite primers generated a total of
  116 polymorphic DNA bands.
• Nei's unbiased measures of genetic identity
  between subpopulations collected during 1997 to
  2001 were conducted.

30
Conclusion The populations were stable over 5
years with a very slight change in genetic
structure.
Very high levels (gt 99) of genetic identity
among the B. dothidea populations collected from
different years in Madera county were observed.
31
D. Pathogen spatial distribution ????????

• Pathogen isolates should be collected form
  different geographical locations.
• Proper molecular makers should be used, and these
  markers should show the polymorphism among
  isolates.
• A PCR with the special primers or probes should
  be performed, and the polymorphic DNA patterns or
  a specific DNA fingerprint can be used in data
  analysis.
• The data analyses may include
• Calculations of genetic diversity and distance
  among geographical populations,
• Clustering analysis to determine the existence of
  isolation or gene flow among geographic
  populations.
• Using UPGMA and genetic identity to determine the
  relationship between genetic distance and
  geographic distances among populations.
• GIS is an useful tool to quantify the above
  relationships,

32
Example 1
Phytopathology 911156-1165

• Isolates were collected from fields with distance
  of few meters to 100 km during the seasons of
  1994 1996.

33

• Each isolate was characterized by
• mating type,
• allozyme at the glucose-6-phosphate isomerase
  (Gpi) and peptidase (Pep) loci,
• restriction fragment length polymorphizm (RFLP)
  with probe RG57 (different races),
• metalaxyl sensitivity and aggressiveness.
• Geostatistics analysis
• Exploratory data analysis.
• Variogram analysis of the spatial structure.
• Surface interpolation.

34

• Six different RFLP genotypes were represented.
• There are two small areas where the three
  genotypes had a similar probability of
  occurrence.
• In 1996-1997, three RFLP banding pattern
  genotypes were found. Genotype B had the
  highest probability of occurrence in most areas
• The genotpe I had a low probability of
  occurrence, with the probability above 0.1 only
  in a small area.

35
E. Pathogen long-distance dispersal and migration
?????????

• Studies focus on origin or long-distance
  dispersal of the plant pathogens.
• Pathogen isolates should be collected from
  different ecological or geographical regions or
  even different continents.
• Different PCR processes could be used by using
  some specific molecular makers including
  species-specific makers, race-specific makers,
  and so forth.
• In addition to normal analysis methods for
  genetic populations, special data analysis
  approaches may include
• Using the parsimony trees to determine the
  distances among populations.
• Testing the hypotheses about mutation, cloning,
  recombination and mating possibility with
  BOOTSTRAP.
• Using the MANTOL method to determine the
  relationship between genetic matrix and
  geographic matrix to conclude about dispersal and
  migration.

36
Example 1
Plant Pathology (2002) 5124-32

1. Isolates from UK, Germany, France and Denmark,
2. Tested Virulence and 16 AFLP genotypes,
3. Used some programs of the software PHYPIL to test
   the null hypotheses A, population is clonal, and
   variation is from mutation, B, migration among
   these countries is not restricted.

37
Isolates were collected from different
countries
38
1. Obtain the length of the parsimony tree of
observed data. 2. Use bootstrap to generate
random populations. 3. Obtain the distribution of
tree lengths of random populations. 4. Compare
observed tree length with the mean tree length of
the distribution.
5. Conclude about the hypotheses.
39
Hypothesis A population is recombining, and
variation is from mutation.
Hypothesis B migration among these countries is
not restricted
Accept the hypothesis B.
Reject the hypothesis A.
40
Science 297537-541.
41
F. Dynamics of pathogen population structures
??????????

• The information on pathogen population structure
  is important to understand pathogen evolution,
  population diversity and related disease
  development.
• Special molecular makers are needed to determine
  the variation of genetic structures of different
  populations and their changes over time and
  space.
• Specific analyses are needed to determine how
  disease development is related to changes in
  pathogen population structures.
• This information is useful to determine disease
  management strategies.

42
Example of Phytothora infestans
From Goodwin, (Phytopathology 87462-473).
43
The distribution of US-1 isolates confirmed with
molecular markers in 19 countries on 4
continents. (From Goodwin, (Phytopathology
87462-473).
44
Distribution of Phytophthora infestans
lineages in Peru. From Perez, et al.
(Phytopathology 91956-965).
45
Distribution of genotypes detected in New
York State in 1992 and 1993.
States and provinces in Northern America, in
which A2 mating type isolates were detected
during 1979-1991. Goodwin et al. (Phytopathology
85473-479).
46
Genotypes of Phytophthora infestans from 5 east
Asian countries
From Koh, et al. (Phytopathology 84922-927)
47
Regions sampled for Phytophthora infestans in The
Netherlands in 1989 and the occurrence of RG57
genotypes by filed
From Drenth, et al. (Phytopathology 831087-1092)
48
Sampling site locations of A1 and A2 isolates of
Phytophthora infestans in Poland from 1985 to
1987 through 1991.
From Sujkowski, et al. (Phytopathology 84201-207)
49
G. Interactions between host resistance and
pathogen virulence??????????

• Studies emphasize on determination of pathogen
  pathogenicity or virulence by using fast and
  accurate molecular methods.
• The important areas may include
• Fast identification of pathogen races,
• Determination of geographic distribution of races
  and different pathogenicities,
• Evaluation of pathogen evolution over time and
  space.
• Providing information on resistance deployment
  and decision support for resistance applications.
• Prediction of disease development and dynamics of
  pathogen races.

50
The ideal system
Pathogenicity pattern
Primer A
Primer B
Race 2- related band
Race 4- related band
Molecular phenotypic pattern
51
Example 1 (type A)

1. Chickpea diseases,
2. RAPD-PCR using 7 OPI primers,
3. 8 Races and 60 isolates,
4. Tested both Pathogenicity and RAPD haplotypes,
5. Analyses using UPGAM and AMOVA.

52
(No Transcript)
53
Isolates of a race
Isolates representing a race
54
The other system

• In this system, the one-by-one relationship
  between a DNA amplicon and a race may be not yet
  found or not exist.
• The group of DNA amplification patterns may be
  needed to distinguish pathogen races.
• Analysis for these cases may include
• Cluster analysis to distinguish populations.
• Genetic distance and identity among populations.
• Comparisons between races by virulence- or
  pathogenicity-tests and DNA phenotypes by
  molecular makers.
• Determination of informative amplicon patterns to
  distinguish races.
• Evaluation of the molecular method for pathogen
  race identification.

55
Example 1 (type B)
Virulence and Molecular Polymorphism in Puccinia
recondita f. sp. tritici in Canada J. A.
Koolmer, J. Q. Liu and M. Sies Phytopathology
85276-285.
56
Basic information

• 19 near-isogenic differential lines,
• RAPD by using 10 arbitrary decamer primers,
• 37 phenotypes were distinguished by the
  differential lines,
• 4. 45 unique phenotypes were distinguished by
  combined virulence and RAPD data,
• 2 major clusters were distinguished with the
  combined data set,
• Data analyses
• Cluster analysis using dissimilarity,
• Comparisons between virulence and molecular
  dissimilarity matrices.

57
Dendrogram based on combined virulence and RAPD
DNA data. 1-5 equivalent grouping of the
isolates. A and B grouping of isolates based on
virulence and molecular polymorphism.
58
Three-dimensional cluster diagram. Number
between clusters represent the average number of
virulence and molecular differences between
isolates in each cluster. Numbers within clusters
represent the average number of virulence and
molecular differences between isolates within the
cluster.
59
Conclusions

• The molecular polymorphisms were more effective
  in distinguishing between the major clusters
  compared to the virulence polymorphisms.
• 2. However, virulence polymorphisms were more
  effective in distinguishing between isolates
  within the major clusters compared to the
  molecular polymorphisms.

60
Book and journals involving molecular epidemiology
61
Software for Population Genetics and Molecular
Epidemiology
Labote, Joanne A. 2000. Software for population
genetic analyses of molecular marker data. Crop
Science 401521-1528.
62
TFPGA
http//herb.bio.nau.edu/miller
Main Functions Diversity
Heterozygosity Expected heterozygosity
Percent polymorphic loci Population structure
F-statistics  G-statistics Homogeneity
Genetic distance Neis  Rogers Pairwise
FST
Equilibrium Hardy-Weinberg
Clustering UPGMA
63
Arlequin
http//anthropologie.unige.ch/arlequin
Main Functions Diversity Heterozygosity
Expected heterozygosity No. alleles/locus
Effective no. alleles Percent polymorphic
loci Population structure F-statistics
ANOVA Rho-statistics Migration
Equilibrium Hardy-Weinberg Two-locus
Genetic distance Neis Pairwise FST
Clustering UPGMA
64
NTSYSpc v2.1

Main Functions Diversity Heterozygosity
Expected heterozygosity No. alleles/locus
Effective no. alleles Percent polymorphic
loci Shannon-Weaver   Equilibrium
Hardy-Weinberg Two-loci Multiloci U-test
Genetic distance Neis   Rogers Pairwise
FST
Clustering Neighbor-Joining UPGMA
Neutrality tests
65
PopGene v3.2
http//www.ualberta.ca/fyeh/index.htm
Main Functions Diversity Heterozygosity
Expected heterozygosity No. alleles/locus
Effective no. alleles Percent polymorphic loci
Shannon-Weaver
Population structure F-statistics  
G-statistics Homogeneity Migration
Equilibrium Hardy-Weinberg Two-loci
Genetic distance Neis Clustering UPGMA
Neutrality tests
66
GDA
http//lewis.eeb.uconn.edu/lewishome/gda.html
Main Functions Diversity Heterozygosity
Expected heterozygosity No. alleles/locus
Percent polymorphic loci Population structure
F-statistics ANOVA
Equilibrium Hardy-Weinberg Two-loci
Multiloci
Genetic distance Neis Pairwise FST
Clustering Neighbor-Joining UPGMA
67
GenePop
http//www.cefe.cnrs-mop.fr
Main Functions Population structure
F-statistics Rho-statistics Homogeneity
Migration Isolation-by-distance   Equilibrium
Hardy-Weinberg Two-loci U-test
68
Phylip v3.6
http//evolution.genetics.washington.edu/phylip.ht
ml
Main Functions Diversity Heterozygosity
Expected heterozygosity No. alleles/locus
Effective no. alleles Percent polymorphic loci
Shannon-Weaver Population structure
F-statistics   G-statistics ANOVA
Genetic distance Neis   Rogers Pairwise
FST
Epuilibrium Hardy-Weinberg Two-locus
Clustering Neighbor-Joining UPGMA
Neutrality tests
69

• ??????????????????? ?
• ?????????????????????????
• ???????????????????????????????????

70
? ?