Nature - USA (2020-09-24)

582 | Nature | Vol 585 | 24 September 2020

Article

that, by definition, this parasite cannot use GYPA for invasion, the
inhibitory impact of Dantu cannot be entirely explained by an altered
PfEBA175–GYPA interaction. Together with our earlier observations that
Dantu decreases invasion both by strains that rely heavily on PfEBA175
(such as Dd2) and by others that are less dependent on this pathway
(3D7), this finding provides further evidence that the impact of Dantu
is not linked to specific receptor–ligand interactions.

Membrane tension and invasion efficiency
To explore the broader biomechanical effects of Dantu on RBCs, we meas-
ured membrane contour fluctuations^15 by combining live imaging with
flickering spectrometry^16 (Extended Data Fig. 5), enabling us to gener-
ate direct measurements of RBC membrane mechanics, such as tension,
radius, viscosity and bending modulus. Tension and bending modulus
represent the resistance of membranes to stretch and bend, respectively.
The bending modulus is affected by lipid composition, membrane thick-
ness and membrane asymmetry, while tension is set by osmotic pressure;
the RBC cytoskeleton can affect both. These phenotypes vary naturally
between RBCs, which can circulate for more than three months after they
are produced from stem cells. As expected because of their lower MCVs,
Dantu RBCs had a significantly higher tension (P = 0.005) and smaller
radius (P = 0.004) than non-Dantu RBCs, but no significant differences
were found in bending modulus (P = 0.469) and viscosity (P = 0.311)
(Fig. 3a). The equatorial radius difference of 0.3 μm between non-Dantu
and Dantu RBCs is minor, and probably explained by a shape change caused

by the increased tension. Tension and radius are linked properties, with

higher tension leading directly to smaller RBCs^17 (Extended Data Fig. 6a).

To test for a link between tension and invasion, we assessed both

parameters simultaneously using video microscopy. We used

high-frame-rate capture to measure tension for all RBCs adjacent to a

rupturing schizont (Supplementary Video 4), and then monitored the

invasion fate of all parasite–RBC contact events with these same RBCs

following schizont rupture (Fig. 3b and Supplementary Video 5). We

observed an intrinsic distribution of membrane tensions in RBCs from

each donor, and discovered a clear association between tension and

invasion: merozoites preferentially invaded nearby RBCs that showed

low tension (Fig. 3c and Supplementary Tables 5, 6). Comparing the dis-

tribution of tension values with invasion efficiency suggested a tension

threshold for successful invasion that was consistent across all three

genotypes. The average of these tension thresholds across the geno-

type groups was 3.8 ( ± 2.0) × 10−7 N m−1, successful invasion being very

rare above this threshold in both Dantu and non-Dantu RBCs (Fig. 3c).

Notably, the median tension was 8.2 × 10−7 N m−1 in Dantu RBCs, meaning

that most were above the tension threshold of 3.8 ( ± 2.0) × 10−7 N m−1

(Fig. 3a). The impact of Dantu on the biomechanical properties of the

RBC is therefore sufficient to explain its impact on invasion, consist-

ent with the invasion-inhibitory effect of Dantu being independent of

a reliance on GYPs or other invasion receptors.

To investigate the mechanism by which higher RBC membrane ten-

sion results in invasion resistance, we examined whether it impedes the

ability of the merozoite to wrap the RBC membrane around itself during

a

b

1

2

4 3

5

**

1.50.40.6 0.81.0 1.2

2.0

2.5

3.0

3.5

4.0

4.5

Parasitaemia (%)

Untreated

0.01

0.001

0.0001

0.00001

0.0005

0.00005

Glutaraldehyde (%)

1.4

Bending modulus (J)

Radius (

μm)

5.0

3.5

4.0

4.5

10 –20

10 –19

Thr

eshold

**

Tension (N m

–1)

Viscosity (Pa s)

0

0.04

0.02

0.06

RBC contours Deformation Invasion Echinocytosis

Non-Dantu

Egress

Dantu heterozygote Dantu homozygote

8 μm

No contactCell 1 No invasionsCells 4,5 InvasionsCells 2,3

e

d

**

Mer

ozoite–RBC contact section (

μm)

Low-tension RBCs

High-tension RBCs

1

2

3

4

** * **

c

Thr

eshold

10 –8

10 –7

10 –6

10 –8

10 –7

10 –6

Tension (N m

–1)

Non-Dantu Dantu homozygote

Successful invasions

Failed invasions

Failed invasions

Successful invasions

Te nsion (N m–1 × 10 –6)

Fig. 3 | Biomechanical properties of the RBC membrane differ across Dantu
genotypes and correlate with invasion. a, Membrane f lickering spectrometry
enabled the measurement and comparison of RBC bending modulus, tension,
radius and viscosity across genotypes (n = 6 individuals per genotype). Means
and standard deviations were obtained from the averages of cell tensions for
each sample, and are as follows: non-Dantu RBCs, (6.0 ± 1.9) × 10−7 N m−1; Dantu
heterozygotes, (7.9 ± 2.8) × 10−7 N m−1; Dantu homozygotes, (8.8 ± 0.7) × 10−7 N m−1.
b, c, We evaluated the impact of tension on parasite invasion by simultaneous
measurement of tension in f lickering analyses and live video imaging of the
invasion process from rupturing schizonts (‘egress’, ‘deformation’, then either
‘invasion’ and ‘echinocytosis’, or a failed invasion) (b), for non-Dantu and Dantu
homozygote RBCs (c). We obtained the threshold range for tension, marked in
a, c, by comparing distributions of tension across Dantu genotypes with their
invasion efficiency. d, We measured the contact region between merozoites

and RBCs, represented in the snapshots at the top, during pre-invasion at the

point of RBC maximum deformation, for low-tension (n = 23) and high-tension

(n = 15) cells (P = 1. 30 × 10−32). The merozoite–RBC contact section was

significantly smaller with high-tension RBCs, meaning that parasites were

much more wrapped around RBCs with a lower membrane tension. e, Parasite

invasion efficiency (parasitaemia) and RBC tension in the presence of six

increasing concentrations of glutaraldehyde (0.00001–0.01%). Parasite

invasion was significantly decreased when RBC tensions were around

0.88 × 10−6 N m−1 (a 22% decrease) and 1.22 × 10−6 N m−1 (a 43% decrease). Median

values are reported from two technical replicates of two biologically

independent samples. Pairwise comparisons between genotypes used the

two-sided Mann–Whitney U-test. **P < 0.01. Numbers of cells and tension are

reported in Supplementary Tables 5, 6.