Title: Global Epidemiology of HIV: Risk Factors, Social Networks and Inter-Subtype Recombinant Strains in Vaccine Trial Sites
1Global Epidemiology of HIV Risk Factors, Social
Networks and Inter-Subtype Recombinant Strains in
Vaccine Trial Sites
FE McCutchan1, GH Kijak1., S Tovanabutra1., E
Sanders-Buell1., C Beyrer2., M deSouza3., M
Arroyo1.,3., M Robb1., D. Birx1., N Michael1.
- US Military HIV Research Program, Rockville, MD,
USA with joint sponsorship by the Division of
AIDS, NIAID, NIH, Bethesda, MD, USA - Johns Hopkins University Bloomberg School of
Public Health, Baltimore, MD, USA - Armed Forces Research Institute of Medical
Sciences, Bangkok, Thailand - Current address Centers for Disease Control
and Prevention, Atlanta, GA, USA
2Background
- The genetic diversity of HIV-1, comprised of
multiple subtypes (clades) and inter-subtype
recombinant forms, poses a significant challenge
to vaccine design and evaluation - The rapid replication rate of HIV-1, coupled with
a high mutation rate and recombination, are the
underlying forces driving diversity - In the infected individual, a swarm of highly
related but non-identical strains, termed a
quasispecies, exists - At the population level, multiple subtypes often
co-circulate, leading to the generation of
inter-subtype recombinant forms - The geographic distribution of subtypes and
recombinant forms is complex and highly uneven
across the globe
3HIV-1 diversity impacts vaccine development at
several stages
- Identification of prevalent strains in the
pandemic and their geographic distribution - Incorporation of prevalent strains into candidate
vaccines - Establishment of vaccine trail cohorts where
prevalent strains circulate, and obtaining a
detailed description of the strains that will
challenge vaccinees - Evaluation of breakthrough infections -
relationship of infecting strains to those in the
vaccine, as a measure of cross-clade immunity
4A complex pandemic
Subtypes A B C D F G H J K
CRF CRF01_AE CRF02_AG - - - - - - CRF35_AD URF
Six Globally Prevalent Strains
5- Current candidate vaccines are based on one or
more of the six globally prevalent strains - Vaccine trials are taking place or planned in
populations where these strains are circulating - Many vaccine trial sites have a complex molecular
epidemiology, with multiple subtypes and
recombinant forms co-circulating
6Vaccine cohorts in East Africa and Asia
CRF01_AE, B, C
A, C, D
7Preparation for vaccine trials in East Africa and
Asia
8HIV-1 Strains in East Africa
Kenya
Uganda
Tanzania
CRF02
120 complete genomes
9Origin and Detection of Recombinant Forms
10HIV-1 is a diploid virus, with two copies of
the RNA genome packaged in the viral particle
- At each replication cycle, RT switches several
times between RNA templates during generation of
the DNA provirus, which is the source of new
viral RNA genomes - Usually, the two RNAs are nearly identical, so
the recombination process does not generate much
new diversity - Some individuals become infected with more than
one subtype of HIV-1, leading to co-packaging of
RNAs from different subtypes into viral
particles now strand switching generates highly
divergent strains with segments from different
subtypes alternating across the genome - Dual infections with more than one subtype are
thought to be the source of inter-subtype
recombinant strains in the pandemic
11A
Detection of Recombination
D
Resolution 300 bp
Bootstrap Value
Position in Genome
12- When more that one HIV-1 subtype circulates in a
population, inter-subtype recombinant forms are
usually present as well - Some recombinant forms begin to spread widely in
populations, and can be recovered from many
individuals widely separated locations - two of
these Circulating recombinant Forms (CRF) are
among the most prevalent globally - Still more recombinants have been recovered only
from a single individual, and are termed Unique
Recombinant Forms (URF) - Little is known about the factors that lead to
the emergence of CRF
13Implications of Recombination for Vaccine Trials
- Dual infection with more than one subtype of
HIV-1 can occur - Participants in vaccine trials may be
challenged simultaneously with multiple
subtypes - Protection from infection or disease progression
may occur when challenged with a single subtype,
but not when challenged with the more diverse
mixture of strains in a dual infected individual - Over time, individuals initially protected from a
single-clade challenge could later succumb to a
multi-clade challenge, obscuring vaccine efficacy - The overall efficacy of vaccines could be blunted
in high risk populations, where dual infections
may be more common, making extrapolation of
results between population sectors with different
risk levels more difficult - The social networks of vaccine trial volunteers
may be important to consider the higher
incidence in high-risk networks may be
counterbalanced by other factors that could
complicate vaccine evaluation
14- For Molecular Epidemiology, recombination poses
new challenges - Need complete sequencing of each HIV-1 strain, or
the use of genotyping tools that evaluate the
subtype in many different regions of each strain - The possibility of dual infection needs to be
accommodated in genotyping approaches - Large populations need to be genotyped to capture
the diversity in mixed-subtype epidemics
15The Epidemiologic Link Between High-risk Groups,
Dual Infection, and Recombinant Strains
16- Fluorescent Multiregion Hybridization Assays for
Regional Application
17Comparative Epidemiology in East Africa
MHAacd assay
MHA
Population High risk females Urban and rural
communities Rural communities Agricultural
Plantation
Country Tanzania Tanzania Uganda Kenya
Cohort HISIS CODE MER Kericho
Genotypes (N) 238 487 329 366 1420
18Comparative Epidemiology in Thailand
MHAbce assay
MHA Genotypes (N) 336 177 293 806
Population Injecting Drug Users Antenatal
Clinic Vaccine Trial Volunteers
Province Chiang Mai Lampang Rayong-Chon Buri
19(No Transcript)
20- High risk cohorts with multiple HIV-1 exposures
have higher rates of dual infection and
recombinant strains - Many different inter-subtype recombinant HIV-1
can emerge in a single dual infected individual,
who may transmit them to others - Many recombinant strains are generated within
high risk social networks, with high rates of
transmission
21Recombinant Strains and Social Networks
22Describing a Recombinant Strain Subtypes and
Breakpoints
1
9200
Subtype A
380
4600
5700
1900
AC Recombinant
1
9200
Subtype C
23- Through recombination, parts of the parental
strains are lost, and cannot be regained until
another dual infection provides opportunity to
recombine again
- Irreversibility lends stability
- Could recombination breakpoints serve as stable
markers through many cycles of transmission,
permitting mapping of the social networks in
which HIV spreads?
380
4600
1900
5700
A
A
A
C
C
Lost Genetic Material
C
C
C
A
A
24Hypotheses
- Mapping of shared breakpoints among recombinant
strains could provide a new dimension to the
molecular epidemiology of HIV-1 - The structure and relationships of recombinant
strains may provide information about the social
networks in which they spread
25Recombinant Strains and Risk Groups
26Recombinant Strains in Low Risk Groups
Transmission
Sampling
single
single
single
single
single
Complete sharing of breakpoints
27Recombinant Strains in High Risk Groups
Transmission
single
dual
single
single
Partial sharing of breakpoints
dual
single
28Complete Genome Sequences of Recombinant Strains
Risk Groups
N
Region
Subtypes
Countries
17 21 26
Uganda Kenya Tanzania
HeterosexualPerinatal
East Africa
A, C, D
42 6 13
Thailand Myanmar China
HeterosexualIDU Unknown
CRF01_AE, B, C
Asia
125
29Fine-mapping of 125 recombinant strains
- Each alignment is 9000 bp
- Inspect alignment base-by-base (600 person-hours)
- Map breakpoint locations onto reference strain
HXB2 - Create database of recombination breakpoints
- A to C at 380
- C to A at 1900
- A to C at 4600
- C to A at 5700
- 896 breakpoints determined
- 521 in strains from East Africa
- 375 in strains from Asia
30- Numerous closely spaced shifts in subtype were
mapped, many of which were too closely spaced to
be efficiently detected and mapped by sliding
window techniques - Overall, the recombinant strains were more
complex than first appreciated
40 of breakpoints are spaced less than 300 nt
apart
Spacing (nt)
Adjacent Breakpoint Pair
31(No Transcript)
32Recombinant HIV Networks and Risk Groups in Asia
- 24 CRF01_AE/B recombinants from Thailand and
Myanmar - 11 from IDU
- 13 from heterosexual transmission
33(No Transcript)
34Expanding the Analysis to 125 Recombinant Strains
from Asia and E. Africa
35Network Visualization Software
- Each strain is a node
- Each shared breakpoint is a connection,
represented by a line - Highly interconnected strains form dense
clusters, with less connected strains at the
periphery
UCiNET and NetDraw by S. Borgatti Boston
College/Analytic Technologies
36(No Transcript)
37- Within these regional epidemics, virtually all of
the recombinant strains analyzed shared at least
one of their breakpoints, and often each of their
breakpoints, with at least one other strain in
the region - Observed a highly connected network of related,
but non-identical recombinants - Degree of networking appeared greater among
Asia vs. E. Africa strains
38How do CRF relate to other strains in their
geographic region?
Criterion 100 bp
CRF07_BC
CRF10_CD
CRF08_BC
CRF34_01B
CRF15_01B
CRF16_A2D
Asia
East Africa
CRF21_A2D
39- CRF share some of their breakpoints with unique
recombinant forms (URF) in their respective
regional epidemics - Distinction between Circulating and Unique
Recombinant forms may be a reflection, not of any
intrinsic property of the virus, but rather a
result of their relative abundance, and
therefore, the probability of re-sampling them,
and their potential for stabilization in
lower-risk networks where dual infection is less
common
40What can be learned about social networks from
the relationships among recombinant strains
circulating within them?
41(No Transcript)
42In Asia
- Heterosexual and IDU Networks in Thailand are
strongly interconnected and these connections
were already established during the first decade
of the Thailand epidemic - Fewer connections across national borders
- Strains from Myanmar bridge China and Thailand
epidemics
43Recombinant Networks in E. Africa
AC
AD
TZ UG KE
CD
ACD A2D
By Subtype
By Country
44In East Africa
- Recombinant strains share some of their
breakpoints even when isolated in different
countries - Given the distances involved, long chains of
transmission undoubtedly separate the sampled
strains, yet breakpoints mark them as part of a
recombinant lineage
45Connections Across National Borders
CD
B/C
AC
AD
Tanzania Uganda Kenya
CRF01_AE/B
China Myanmar Thailand
A2D
Asia
E. Africa
46- Highly interconnected networks of related strains
were observed, both in E. Africa and in Asia - Networks of recombinant strains were more
tightly interconnected in higher risk groups,
like IDU, than in general population cohorts - Dual infection in high-risk groups, which is
required to generate the recombinant strain
networks that we observe is occurring to a
considerable degree - CRF may represent recombinant strains that have
entered much lower-risk networks, where the
opportunity for dual infection and further
recombination is limited - reproductive isolation
47Conclusion
- HIV-1 strains reflect the social networks in
which they are spreading - Recombinant strains can be useful for delineating
the boundaries and dynamics of the social
networks in which they spread - High-risk groups may contribute
disproportionately to the overall diversity of
the pandemic, because of increased dual infection
and recombination - Testing candidate vaccines in many types of
social networks, with different levels of risk,
may be an optimal strategy in the quest for an
HIV-1 vaccine
48Contributors
- Participants in cohort development and other
studies in Tanzania, Uganda, Kenya, Thailand,
China, Myanmar - Oliver Hoffmann, Steffan Geis, Leonard Maboko,
Donan Mmbando, Eluter Samky, Michael Hoelscher,
and other members of the Mbeya Medical Research
Programme, Tanzania - David Serwadda, Nelson Sewankambo, Maria Wawer,
Ron Gray Makerere University and Uganda Virus
Research Institute, Uganda,Columbia University
and Johns Hopkins University, USA and other
members of the Rakai Project, Uganda - Carl Mason, USAMRU-K, Monique Wassuna, KEMRI, and
other contributors to the Kenya Blood Bank Study - David Celentano, Chris Beyrer, Vinai Suriyanon,
Jaroon Jittiwutikarn, Thira Sirisanthana, Myat
Htoo Razak and other contributors to the Opiate
Users Research Study, Thailand - Vilaiwan Gulgolgarn, Manu Wera-arpachai, Chirasak
Khamboonrueng, Kenrad Nelson, Nakorn Dabbhasuta
and others from the Lampang perinatal
transmission cohort study, Thailand - Supachai Rerks-Ngarm, Sonchai Wattana, Wiwat
Wiriyakijja, Sorachai Nitayaphan, Chirapa
Eamsila, Jerome Kim, Michael Benenson, Arthur
Brown and others for samples from volunteers
deferred from enrollment in the Phase III
prime-boost vaccine trial in Rayong-Chonburi
Provinces, Thailand - Special thanks to Jocelyn Chiu and her mentors,
Sodsai Tovanabutra, and Eric Sanders-Buell, for
inspection and analysis of 1,125,000 nucleotides
of sequence alignment